fix: honor parser params and image VLM system_prompt in Go ingestion (#17334)
## Summary
Fix the Go ingestion pipeline so that several parser setup switches and
the
image VLM prompt are actually honored end-to-end (previously the DSL
fields
existed but the Go code never read them).
- **DOCX** (`docx_parser.go`, `docx_postprocess.go`): read `remove_toc`
and
`remove_header_footer`; apply to both JSON and markdown output paths
(outline-based TOC removal with a text-heuristic fallback, plus
header/footer section filtering).
- **HTML** (`html_parser.go`, `html_postprocess.go`, `text_toc.go`):
read
`remove_header_footer` (pre-parse strip of `<header>`/`<footer>` and
ARIA
`banner`/`contentinfo`) and `remove_toc` (post-parse
`remove_contents_table`
heuristic).
- **Markdown** (`markdown_parser.go`): read `flatten_media_to_text` and
force
media blocks to text when enabled.
- **Image VLM** (`media_dispatch.go`): read `system_prompt` instead of
`prompt`
so the user-configured image VLM prompt is no longer silently dropped
(`prompt` remains the video family key).
All flags are wired through `ConfigureFromSetup`, which the dispatch
layer
already invokes for every family, so the behavior is live rather than
dead code.
## Test plan
- New unit tests: `docx_postprocess_test.go`, `html_parser_test.go`,
`text_toc_test.go`, `markdown_parser_test.go`, `media_dispatch_test.go`.
- `bash build.sh --test ./internal/parser/parser/...
./internal/ingestion/component/...`
## Notes
- The `File` component is excluded from this migration scope.
- Relates to the Python→Go parity diff (Parser 1.8–1.11, 1.15).
2026-07-24 14:42:26 +08:00
|
|
|
//
|
|
|
|
|
// Copyright 2026 The InfiniFlow Authors. All Rights Reserved.
|
|
|
|
|
//
|
|
|
|
|
// Licensed under the Apache License, Version 2.0 (the "License");
|
|
|
|
|
// you may not use this file except in compliance with the License.
|
|
|
|
|
// You may obtain a copy of the License at
|
|
|
|
|
//
|
|
|
|
|
// http://www.apache.org/licenses/LICENSE-2.0
|
|
|
|
|
//
|
|
|
|
|
// Unless required by applicable law or agreed to in writing, software
|
|
|
|
|
// distributed under the License is distributed on an "AS IS" BASIS,
|
|
|
|
|
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
|
|
|
// See the License for the specific language governing permissions and
|
|
|
|
|
// limitations under the License.
|
|
|
|
|
|
|
|
|
|
package component
|
|
|
|
|
|
|
|
|
|
import (
|
|
|
|
|
"context"
|
2026-07-27 19:30:41 +08:00
|
|
|
"ragflow/internal/dao"
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
"strings"
|
fix: honor parser params and image VLM system_prompt in Go ingestion (#17334)
## Summary
Fix the Go ingestion pipeline so that several parser setup switches and
the
image VLM prompt are actually honored end-to-end (previously the DSL
fields
existed but the Go code never read them).
- **DOCX** (`docx_parser.go`, `docx_postprocess.go`): read `remove_toc`
and
`remove_header_footer`; apply to both JSON and markdown output paths
(outline-based TOC removal with a text-heuristic fallback, plus
header/footer section filtering).
- **HTML** (`html_parser.go`, `html_postprocess.go`, `text_toc.go`):
read
`remove_header_footer` (pre-parse strip of `<header>`/`<footer>` and
ARIA
`banner`/`contentinfo`) and `remove_toc` (post-parse
`remove_contents_table`
heuristic).
- **Markdown** (`markdown_parser.go`): read `flatten_media_to_text` and
force
media blocks to text when enabled.
- **Image VLM** (`media_dispatch.go`): read `system_prompt` instead of
`prompt`
so the user-configured image VLM prompt is no longer silently dropped
(`prompt` remains the video family key).
All flags are wired through `ConfigureFromSetup`, which the dispatch
layer
already invokes for every family, so the behavior is live rather than
dead code.
## Test plan
- New unit tests: `docx_postprocess_test.go`, `html_parser_test.go`,
`text_toc_test.go`, `markdown_parser_test.go`, `media_dispatch_test.go`.
- `bash build.sh --test ./internal/parser/parser/...
./internal/ingestion/component/...`
## Notes
- The `File` component is excluded from this migration scope.
- Relates to the Python→Go parity diff (Parser 1.8–1.11, 1.15).
2026-07-24 14:42:26 +08:00
|
|
|
"sync"
|
|
|
|
|
"testing"
|
|
|
|
|
|
|
|
|
|
"ragflow/internal/common"
|
|
|
|
|
"ragflow/internal/entity"
|
|
|
|
|
modelModule "ragflow/internal/entity/models"
|
|
|
|
|
"ragflow/internal/utility"
|
2026-07-27 19:30:41 +08:00
|
|
|
|
|
|
|
|
"gorm.io/gorm"
|
fix: honor parser params and image VLM system_prompt in Go ingestion (#17334)
## Summary
Fix the Go ingestion pipeline so that several parser setup switches and
the
image VLM prompt are actually honored end-to-end (previously the DSL
fields
existed but the Go code never read them).
- **DOCX** (`docx_parser.go`, `docx_postprocess.go`): read `remove_toc`
and
`remove_header_footer`; apply to both JSON and markdown output paths
(outline-based TOC removal with a text-heuristic fallback, plus
header/footer section filtering).
- **HTML** (`html_parser.go`, `html_postprocess.go`, `text_toc.go`):
read
`remove_header_footer` (pre-parse strip of `<header>`/`<footer>` and
ARIA
`banner`/`contentinfo`) and `remove_toc` (post-parse
`remove_contents_table`
heuristic).
- **Markdown** (`markdown_parser.go`): read `flatten_media_to_text` and
force
media blocks to text when enabled.
- **Image VLM** (`media_dispatch.go`): read `system_prompt` instead of
`prompt`
so the user-configured image VLM prompt is no longer silently dropped
(`prompt` remains the video family key).
All flags are wired through `ConfigureFromSetup`, which the dispatch
layer
already invokes for every family, so the behavior is live rather than
dead code.
## Test plan
- New unit tests: `docx_postprocess_test.go`, `html_parser_test.go`,
`text_toc_test.go`, `markdown_parser_test.go`, `media_dispatch_test.go`.
- `bash build.sh --test ./internal/parser/parser/...
./internal/ingestion/component/...`
## Notes
- The `File` component is excluded from this migration scope.
- Relates to the Python→Go parity diff (Parser 1.8–1.11, 1.15).
2026-07-24 14:42:26 +08:00
|
|
|
)
|
|
|
|
|
|
|
|
|
|
// imagePromptCaptureDriver embeds ModelDriver so it satisfies the interface
|
|
|
|
|
// without listing every method; only ChatWithMessages is overridden to record
|
|
|
|
|
// the messages the image branch selected.
|
|
|
|
|
type imagePromptCaptureDriver struct {
|
|
|
|
|
modelModule.ModelDriver
|
|
|
|
|
mu sync.Mutex
|
|
|
|
|
captured []modelModule.Message
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
func (d *imagePromptCaptureDriver) ChatWithMessages(ctx context.Context, modelName string, messages []modelModule.Message, apiConfig *modelModule.APIConfig, chatModelConfig *modelModule.ChatConfig, usage *common.ModelUsage) (*modelModule.ChatResponse, error) {
|
|
|
|
|
d.mu.Lock()
|
|
|
|
|
d.captured = append(d.captured, messages...)
|
|
|
|
|
d.mu.Unlock()
|
|
|
|
|
ans := "captured"
|
|
|
|
|
return &modelModule.ChatResponse{Answer: &ans}, nil
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// firstUserText extracts the text of the first "text" content part from the
|
|
|
|
|
// first captured user message. It scans parts by the "type" discriminator so
|
|
|
|
|
// the test stays valid regardless of part ordering (image_url may precede text).
|
|
|
|
|
func firstUserText(msgs []modelModule.Message) (string, bool) {
|
|
|
|
|
if len(msgs) == 0 {
|
|
|
|
|
return "", false
|
|
|
|
|
}
|
|
|
|
|
parts, ok := msgs[0].Content.([]interface{})
|
|
|
|
|
if !ok || len(parts) == 0 {
|
|
|
|
|
return "", false
|
|
|
|
|
}
|
|
|
|
|
for _, raw := range parts {
|
|
|
|
|
part, ok := raw.(map[string]any)
|
|
|
|
|
if !ok {
|
|
|
|
|
continue
|
|
|
|
|
}
|
|
|
|
|
if part["type"] != "text" {
|
|
|
|
|
continue
|
|
|
|
|
}
|
|
|
|
|
if txt, ok := part["text"].(string); ok {
|
|
|
|
|
return txt, true
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return "", false
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// TestMaybeDispatchImage_UsesSystemPrompt pins that the image branch reads the
|
|
|
|
|
// `system_prompt` setup key (Python/DSL contract), not `prompt`. Before the fix
|
|
|
|
|
// the branch read setup["prompt"] which is empty for image, so the VLM received
|
|
|
|
|
// the hardcoded default instead of the user-configured value.
|
|
|
|
|
func TestMaybeDispatchImage_UsesSystemPrompt(t *testing.T) {
|
|
|
|
|
origResolver := resolveTenantModelByType
|
|
|
|
|
defer func() { resolveTenantModelByType = origResolver }()
|
|
|
|
|
|
|
|
|
|
drv := &imagePromptCaptureDriver{}
|
2026-07-27 19:30:41 +08:00
|
|
|
resolveTenantModelByType = func(ctx context.Context, db *gorm.DB, tenantID string, modelType entity.ModelType) (modelModule.ModelDriver, string, *modelModule.APIConfig, int, error) {
|
fix: honor parser params and image VLM system_prompt in Go ingestion (#17334)
## Summary
Fix the Go ingestion pipeline so that several parser setup switches and
the
image VLM prompt are actually honored end-to-end (previously the DSL
fields
existed but the Go code never read them).
- **DOCX** (`docx_parser.go`, `docx_postprocess.go`): read `remove_toc`
and
`remove_header_footer`; apply to both JSON and markdown output paths
(outline-based TOC removal with a text-heuristic fallback, plus
header/footer section filtering).
- **HTML** (`html_parser.go`, `html_postprocess.go`, `text_toc.go`):
read
`remove_header_footer` (pre-parse strip of `<header>`/`<footer>` and
ARIA
`banner`/`contentinfo`) and `remove_toc` (post-parse
`remove_contents_table`
heuristic).
- **Markdown** (`markdown_parser.go`): read `flatten_media_to_text` and
force
media blocks to text when enabled.
- **Image VLM** (`media_dispatch.go`): read `system_prompt` instead of
`prompt`
so the user-configured image VLM prompt is no longer silently dropped
(`prompt` remains the video family key).
All flags are wired through `ConfigureFromSetup`, which the dispatch
layer
already invokes for every family, so the behavior is live rather than
dead code.
## Test plan
- New unit tests: `docx_postprocess_test.go`, `html_parser_test.go`,
`text_toc_test.go`, `markdown_parser_test.go`, `media_dispatch_test.go`.
- `bash build.sh --test ./internal/parser/parser/...
./internal/ingestion/component/...`
## Notes
- The `File` component is excluded from this migration scope.
- Relates to the Python→Go parity diff (Parser 1.8–1.11, 1.15).
2026-07-24 14:42:26 +08:00
|
|
|
return drv, "img-model", &modelModule.APIConfig{}, 0, nil
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
setups := defaultSetups()
|
|
|
|
|
// image family's contract key is system_prompt (parser.go:295).
|
|
|
|
|
// Also set a legacy `prompt` sentinel to assert system_prompt wins
|
|
|
|
|
// when both keys are present (regression guard against re-reading
|
|
|
|
|
// setup["prompt"], which is the video-family key, not image).
|
|
|
|
|
setups["image"]["prompt"] = "legacy prompt"
|
|
|
|
|
setups["image"]["system_prompt"] = "自定义视觉提示"
|
|
|
|
|
|
2026-07-27 19:30:41 +08:00
|
|
|
ctx := t.Context()
|
fix: honor parser params and image VLM system_prompt in Go ingestion (#17334)
## Summary
Fix the Go ingestion pipeline so that several parser setup switches and
the
image VLM prompt are actually honored end-to-end (previously the DSL
fields
existed but the Go code never read them).
- **DOCX** (`docx_parser.go`, `docx_postprocess.go`): read `remove_toc`
and
`remove_header_footer`; apply to both JSON and markdown output paths
(outline-based TOC removal with a text-heuristic fallback, plus
header/footer section filtering).
- **HTML** (`html_parser.go`, `html_postprocess.go`, `text_toc.go`):
read
`remove_header_footer` (pre-parse strip of `<header>`/`<footer>` and
ARIA
`banner`/`contentinfo`) and `remove_toc` (post-parse
`remove_contents_table`
heuristic).
- **Markdown** (`markdown_parser.go`): read `flatten_media_to_text` and
force
media blocks to text when enabled.
- **Image VLM** (`media_dispatch.go`): read `system_prompt` instead of
`prompt`
so the user-configured image VLM prompt is no longer silently dropped
(`prompt` remains the video family key).
All flags are wired through `ConfigureFromSetup`, which the dispatch
layer
already invokes for every family, so the behavior is live rather than
dead code.
## Test plan
- New unit tests: `docx_postprocess_test.go`, `html_parser_test.go`,
`text_toc_test.go`, `markdown_parser_test.go`, `media_dispatch_test.go`.
- `bash build.sh --test ./internal/parser/parser/...
./internal/ingestion/component/...`
## Notes
- The `File` component is excluded from this migration scope.
- Relates to the Python→Go parity diff (Parser 1.8–1.11, 1.15).
2026-07-24 14:42:26 +08:00
|
|
|
res, dispatched, err := maybeDispatchImage(
|
2026-07-27 19:30:41 +08:00
|
|
|
ctx,
|
|
|
|
|
dao.DB,
|
fix: honor parser params and image VLM system_prompt in Go ingestion (#17334)
## Summary
Fix the Go ingestion pipeline so that several parser setup switches and
the
image VLM prompt are actually honored end-to-end (previously the DSL
fields
existed but the Go code never read them).
- **DOCX** (`docx_parser.go`, `docx_postprocess.go`): read `remove_toc`
and
`remove_header_footer`; apply to both JSON and markdown output paths
(outline-based TOC removal with a text-heuristic fallback, plus
header/footer section filtering).
- **HTML** (`html_parser.go`, `html_postprocess.go`, `text_toc.go`):
read
`remove_header_footer` (pre-parse strip of `<header>`/`<footer>` and
ARIA
`banner`/`contentinfo`) and `remove_toc` (post-parse
`remove_contents_table`
heuristic).
- **Markdown** (`markdown_parser.go`): read `flatten_media_to_text` and
force
media blocks to text when enabled.
- **Image VLM** (`media_dispatch.go`): read `system_prompt` instead of
`prompt`
so the user-configured image VLM prompt is no longer silently dropped
(`prompt` remains the video family key).
All flags are wired through `ConfigureFromSetup`, which the dispatch
layer
already invokes for every family, so the behavior is live rather than
dead code.
## Test plan
- New unit tests: `docx_postprocess_test.go`, `html_parser_test.go`,
`text_toc_test.go`, `markdown_parser_test.go`, `media_dispatch_test.go`.
- `bash build.sh --test ./internal/parser/parser/...
./internal/ingestion/component/...`
## Notes
- The `File` component is excluded from this migration scope.
- Relates to the Python→Go parity diff (Parser 1.8–1.11, 1.15).
2026-07-24 14:42:26 +08:00
|
|
|
utility.FileTypeVISUAL,
|
|
|
|
|
"test.png",
|
|
|
|
|
[]byte("not-a-real-image"),
|
|
|
|
|
map[string]any{"tenant_id": "t1"},
|
|
|
|
|
setups,
|
|
|
|
|
)
|
|
|
|
|
if err != nil {
|
|
|
|
|
t.Fatalf("maybeDispatchImage: %v", err)
|
|
|
|
|
}
|
|
|
|
|
if !dispatched {
|
|
|
|
|
t.Fatalf("expected dispatched=true for VISUAL file")
|
|
|
|
|
}
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
// After the output-shape fix the image branch returns JSON items
|
|
|
|
|
// (OutputFormat=="json"), not a bare Text field. The combined text
|
|
|
|
|
// now lives in JSON[0]["text"]; the legacy res.Text is no longer
|
|
|
|
|
// populated for the image family.
|
|
|
|
|
if res.OutputFormat != "json" {
|
|
|
|
|
t.Fatalf("OutputFormat = %q, want json (image family is always structured)", res.OutputFormat)
|
|
|
|
|
}
|
|
|
|
|
if len(res.JSON) != 1 {
|
|
|
|
|
t.Fatalf("JSON len = %d, want 1 (image result must be a single JSON item)", len(res.JSON))
|
|
|
|
|
}
|
|
|
|
|
if txt, _ := res.JSON[0]["text"].(string); txt == "" {
|
|
|
|
|
t.Fatalf("expected non-empty combined text in JSON[0][\"text\"]")
|
fix: honor parser params and image VLM system_prompt in Go ingestion (#17334)
## Summary
Fix the Go ingestion pipeline so that several parser setup switches and
the
image VLM prompt are actually honored end-to-end (previously the DSL
fields
existed but the Go code never read them).
- **DOCX** (`docx_parser.go`, `docx_postprocess.go`): read `remove_toc`
and
`remove_header_footer`; apply to both JSON and markdown output paths
(outline-based TOC removal with a text-heuristic fallback, plus
header/footer section filtering).
- **HTML** (`html_parser.go`, `html_postprocess.go`, `text_toc.go`):
read
`remove_header_footer` (pre-parse strip of `<header>`/`<footer>` and
ARIA
`banner`/`contentinfo`) and `remove_toc` (post-parse
`remove_contents_table`
heuristic).
- **Markdown** (`markdown_parser.go`): read `flatten_media_to_text` and
force
media blocks to text when enabled.
- **Image VLM** (`media_dispatch.go`): read `system_prompt` instead of
`prompt`
so the user-configured image VLM prompt is no longer silently dropped
(`prompt` remains the video family key).
All flags are wired through `ConfigureFromSetup`, which the dispatch
layer
already invokes for every family, so the behavior is live rather than
dead code.
## Test plan
- New unit tests: `docx_postprocess_test.go`, `html_parser_test.go`,
`text_toc_test.go`, `markdown_parser_test.go`, `media_dispatch_test.go`.
- `bash build.sh --test ./internal/parser/parser/...
./internal/ingestion/component/...`
## Notes
- The `File` component is excluded from this migration scope.
- Relates to the Python→Go parity diff (Parser 1.8–1.11, 1.15).
2026-07-24 14:42:26 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
got, ok := firstUserText(drv.captured)
|
|
|
|
|
if !ok {
|
|
|
|
|
t.Fatalf("no user text captured in VLM messages: %#v", drv.captured)
|
|
|
|
|
}
|
|
|
|
|
if got != "自定义视觉提示" {
|
|
|
|
|
t.Fatalf("VLM user text = %q, want %q (image branch must read system_prompt)", got, "自定义视觉提示")
|
|
|
|
|
}
|
|
|
|
|
}
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
|
|
|
|
|
// TestMaybeDispatchImage_ReturnsJSONWithImage pins the output-shape fix:
|
|
|
|
|
// the image branch must return a JSON item carrying the `image` attachment
|
|
|
|
|
// (data URI) and `doc_type_kwd:"image"`, mirroring Python
|
|
|
|
|
// rag/app/picture.py:71-72. Before the fix the branch returned a bare Text
|
|
|
|
|
// string with JSON=nil, dropping the image attachment and (on the default
|
|
|
|
|
// json path) causing OneChunker/TokenChunker to reject the payload.
|
|
|
|
|
func TestMaybeDispatchImage_ReturnsJSONWithImage(t *testing.T) {
|
|
|
|
|
origResolver := resolveTenantModelByType
|
|
|
|
|
defer func() { resolveTenantModelByType = origResolver }()
|
|
|
|
|
|
|
|
|
|
drv := &imagePromptCaptureDriver{}
|
2026-07-27 19:30:41 +08:00
|
|
|
resolveTenantModelByType = func(ctx context.Context, db *gorm.DB, tenantID string, modelType entity.ModelType) (modelModule.ModelDriver, string, *modelModule.APIConfig, int, error) {
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
return drv, "img-model", &modelModule.APIConfig{}, 0, nil
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
setups := defaultSetups()
|
2026-07-27 19:30:41 +08:00
|
|
|
ctx := t.Context()
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
res, dispatched, err := maybeDispatchImage(
|
2026-07-27 19:30:41 +08:00
|
|
|
ctx,
|
|
|
|
|
dao.DB,
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
utility.FileTypeVISUAL,
|
|
|
|
|
"test.png",
|
|
|
|
|
[]byte("not-a-real-image"),
|
|
|
|
|
map[string]any{"tenant_id": "t1"},
|
|
|
|
|
setups,
|
|
|
|
|
)
|
|
|
|
|
if err != nil {
|
|
|
|
|
t.Fatalf("maybeDispatchImage: %v", err)
|
|
|
|
|
}
|
|
|
|
|
if !dispatched {
|
|
|
|
|
t.Fatalf("expected dispatched=true")
|
|
|
|
|
}
|
|
|
|
|
if res.OutputFormat != "json" {
|
|
|
|
|
t.Fatalf("OutputFormat = %q, want json", res.OutputFormat)
|
|
|
|
|
}
|
|
|
|
|
if len(res.JSON) != 1 {
|
|
|
|
|
t.Fatalf("JSON len = %d, want 1", len(res.JSON))
|
|
|
|
|
}
|
|
|
|
|
item := res.JSON[0]
|
|
|
|
|
if got, _ := item["doc_type_kwd"].(string); got != "image" {
|
|
|
|
|
t.Errorf("doc_type_kwd = %q, want \"image\"", got)
|
|
|
|
|
}
|
|
|
|
|
img, _ := item["image"].(string)
|
|
|
|
|
if !strings.HasPrefix(img, "data:") || !strings.Contains(img, ";base64,") {
|
|
|
|
|
t.Errorf("image = %q, want a data URI (data:<mime>;base64,<b64>)", img)
|
|
|
|
|
}
|
|
|
|
|
if txt, _ := item["text"].(string); txt == "" {
|
|
|
|
|
t.Errorf("text field empty; want non-empty combined OCR+VLM text")
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// TestMaybeDispatchImage_HardcodesJSONOutput verifies the image family
|
|
|
|
|
// always emits json regardless of setup["output_format"]. Python
|
|
|
|
|
// rag/app/picture.py:chunk() has no output_format concept — it always
|
|
|
|
|
// returns a structured doc. Honoring a "text" override produced a bare
|
|
|
|
|
// Text payload that lost the image attachment and set doc_type to "text".
|
|
|
|
|
func TestMaybeDispatchImage_HardcodesJSONOutput(t *testing.T) {
|
|
|
|
|
origResolver := resolveTenantModelByType
|
|
|
|
|
defer func() { resolveTenantModelByType = origResolver }()
|
|
|
|
|
|
|
|
|
|
drv := &imagePromptCaptureDriver{}
|
2026-07-27 19:30:41 +08:00
|
|
|
resolveTenantModelByType = func(ctx context.Context, db *gorm.DB, tenantID string, modelType entity.ModelType) (modelModule.ModelDriver, string, *modelModule.APIConfig, int, error) {
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
return drv, "img-model", &modelModule.APIConfig{}, 0, nil
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
setups := defaultSetups()
|
|
|
|
|
setups["image"]["output_format"] = "text" // legacy/override; must be ignored
|
2026-07-27 19:30:41 +08:00
|
|
|
ctx := t.Context()
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
res, _, err := maybeDispatchImage(
|
2026-07-27 19:30:41 +08:00
|
|
|
ctx,
|
|
|
|
|
dao.DB,
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
utility.FileTypeVISUAL,
|
|
|
|
|
"test.png",
|
|
|
|
|
[]byte("not-a-real-image"),
|
|
|
|
|
map[string]any{"tenant_id": "t1"},
|
|
|
|
|
setups,
|
|
|
|
|
)
|
|
|
|
|
if err != nil {
|
|
|
|
|
t.Fatalf("maybeDispatchImage: %v", err)
|
|
|
|
|
}
|
|
|
|
|
if res.OutputFormat != "json" {
|
|
|
|
|
t.Fatalf("OutputFormat = %q, want json (image family must ignore output_format override)", res.OutputFormat)
|
|
|
|
|
}
|
|
|
|
|
if len(res.JSON) != 1 {
|
|
|
|
|
t.Fatalf("JSON len = %d, want 1 even when setup says text", len(res.JSON))
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// audioTranscribeDriver is a mock ModelDriver whose TranscribeAudio returns a
|
|
|
|
|
// fixed transcription, so maybeDispatchAudio can be exercised without a real
|
|
|
|
|
// ASR provider.
|
|
|
|
|
type audioTranscribeDriver struct {
|
|
|
|
|
modelModule.ModelDriver
|
|
|
|
|
transcription string
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
func (d *audioTranscribeDriver) TranscribeAudio(ctx context.Context, _ *string, _ *string, _ *modelModule.APIConfig, _ *modelModule.ASRConfig, _ *common.ModelUsage) (*modelModule.ASRResponse, error) {
|
|
|
|
|
return &modelModule.ASRResponse{Text: d.transcription}, nil
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// TestMaybeDispatchAudio_JSONCarriesTranscription pins diff 2.11: when the
|
|
|
|
|
// audio family's output_format is "json", the ASR transcription must be
|
|
|
|
|
// carried in the JSON items (not only in the Text field). Before the fix the
|
|
|
|
|
// branch returned Text only with an empty JSON slice, and the Invoke switch
|
|
|
|
|
// silently dropped the transcription because it has no "json" branch.
|
|
|
|
|
func TestMaybeDispatchAudio_JSONCarriesTranscription(t *testing.T) {
|
|
|
|
|
origResolver := resolveTenantModelByType
|
|
|
|
|
defer func() { resolveTenantModelByType = origResolver }()
|
|
|
|
|
|
|
|
|
|
const want = "hello world"
|
|
|
|
|
drv := &audioTranscribeDriver{transcription: want}
|
2026-07-27 19:30:41 +08:00
|
|
|
resolveTenantModelByType = func(ctx context.Context, db *gorm.DB, tenantID string, modelType entity.ModelType) (modelModule.ModelDriver, string, *modelModule.APIConfig, int, error) {
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
return drv, "asr-model", &modelModule.APIConfig{}, 0, nil
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
setups := defaultSetups()
|
|
|
|
|
setups["audio"]["output_format"] = "json"
|
|
|
|
|
|
2026-07-27 19:30:41 +08:00
|
|
|
ctx := t.Context()
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
res, dispatched, err := maybeDispatchAudio(
|
2026-07-27 19:30:41 +08:00
|
|
|
ctx,
|
|
|
|
|
dao.DB,
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
utility.FileTypeAURAL,
|
|
|
|
|
"test.mp3",
|
|
|
|
|
[]byte("fake-audio"),
|
|
|
|
|
map[string]any{"tenant_id": "t1"},
|
|
|
|
|
setups,
|
|
|
|
|
)
|
|
|
|
|
if err != nil {
|
|
|
|
|
t.Fatalf("maybeDispatchAudio: %v", err)
|
|
|
|
|
}
|
|
|
|
|
if !dispatched {
|
|
|
|
|
t.Fatalf("expected dispatched=true for AURAL file")
|
|
|
|
|
}
|
|
|
|
|
if res.OutputFormat != "json" {
|
|
|
|
|
t.Fatalf("OutputFormat = %q, want json", res.OutputFormat)
|
|
|
|
|
}
|
|
|
|
|
if len(res.JSON) != 1 {
|
|
|
|
|
t.Fatalf("JSON len = %d, want 1 (transcription must be carried as a JSON item)", len(res.JSON))
|
|
|
|
|
}
|
|
|
|
|
if got, _ := res.JSON[0]["text"].(string); got != want {
|
|
|
|
|
t.Fatalf("JSON[0].text = %q, want %q", got, want)
|
|
|
|
|
}
|
|
|
|
|
if got, _ := res.JSON[0]["doc_type_kwd"].(string); got != "audio" {
|
|
|
|
|
t.Fatalf("JSON[0].doc_type_kwd = %q, want audio", got)
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// TestMaybeDispatchAudio_TextCarriesTranscription guards the text path: with
|
|
|
|
|
// output_format "text" the transcription stays in the Text field and JSON is
|
|
|
|
|
// empty (current default after aligning with Python parser.py:232).
|
|
|
|
|
func TestMaybeDispatchAudio_TextCarriesTranscription(t *testing.T) {
|
|
|
|
|
origResolver := resolveTenantModelByType
|
|
|
|
|
defer func() { resolveTenantModelByType = origResolver }()
|
|
|
|
|
|
|
|
|
|
const want = "hello world"
|
|
|
|
|
drv := &audioTranscribeDriver{transcription: want}
|
2026-07-27 19:30:41 +08:00
|
|
|
resolveTenantModelByType = func(ctx context.Context, db *gorm.DB, tenantID string, modelType entity.ModelType) (modelModule.ModelDriver, string, *modelModule.APIConfig, int, error) {
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
return drv, "asr-model", &modelModule.APIConfig{}, 0, nil
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
setups := defaultSetups()
|
|
|
|
|
setups["audio"]["output_format"] = "text"
|
|
|
|
|
|
2026-07-27 19:30:41 +08:00
|
|
|
ctx := t.Context()
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
res, dispatched, err := maybeDispatchAudio(
|
2026-07-27 19:30:41 +08:00
|
|
|
ctx,
|
|
|
|
|
dao.DB,
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
utility.FileTypeAURAL,
|
|
|
|
|
"test.mp3",
|
|
|
|
|
[]byte("fake-audio"),
|
|
|
|
|
map[string]any{"tenant_id": "t1"},
|
|
|
|
|
setups,
|
|
|
|
|
)
|
|
|
|
|
if err != nil {
|
|
|
|
|
t.Fatalf("maybeDispatchAudio: %v", err)
|
|
|
|
|
}
|
|
|
|
|
if !dispatched {
|
|
|
|
|
t.Fatalf("expected dispatched=true for AURAL file")
|
|
|
|
|
}
|
|
|
|
|
if res.OutputFormat != "text" {
|
|
|
|
|
t.Fatalf("OutputFormat = %q, want text", res.OutputFormat)
|
|
|
|
|
}
|
|
|
|
|
if res.Text != want {
|
|
|
|
|
t.Fatalf("Text = %q, want %q", res.Text, want)
|
|
|
|
|
}
|
|
|
|
|
if len(res.JSON) != 0 {
|
|
|
|
|
t.Fatalf("JSON len = %d, want 0 for text output", len(res.JSON))
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
fix: Go ingestion migration batch 5 (Parser 1.1/1.7/2.11, Chunker 1.7/1.8/2.6/2.7, Tokenizer 6x fixes) (#17419)
## Summary
Continuation of the Python→Go ingestion pipeline migration (File →
Parser → Chunker → Extractor → Tokenizer). Fixes cover Parser, Chunker,
and Tokenizer gaps identified. Fix page number (0-indexed and 1-index
mixed before fix; use 1-indexed after fix) and chunk order issues.
### Parser
- **Slides TCADP (1.7):** `pptx_tcadp.go` + TCADP branch in
`pptx_parser.go`/`ppt_parser.go` — PowerPoint files now support
`parse_method="tcadp"` via the TCADP cloud service, matching the
spreadsheet-family TCADP pattern. PPT containers pass `"PPT"` as
fileType (not hardcoded `"PPTX"`).
- **Audio default output_format (2.11):** `defaultSetups()` audio
default changed from `"text"` to `"json"`, aligning with Python
`parser.py:232` and `AllowedOutputFormat["audio"]={"json"}`.
- **PDF VLM enhancement (1.1):** `maybeDispatchPDFVisionEnhancement` in
`pdf_vision_dispatch.go` enriches image/table items with IMAGE2TEXT
model descriptions after PDF parsing, mirroring Python
`enhance_media_sections_with_vision`. Semaphore fix: acquire before
goroutine start to prevent unbounded goroutine creation.
- **json family (2.3):** reclassified as Keep Go — `json_parser.go` is a
functional enhancement, not a parity gap.
- **page number:** changed from "mixed use of 1-indexed & 0-indexed" to
"1-indexed"
### Chunker
- **BULLET_PATTERN fallback (1.7):** 4th-level fallback in
`resolveTitleLevels` (`title.go`) detects bullet/numbered-list patterns
(Chinese legal, numbering, English) when outline + regex levels produce
only bodyLevel. Guarded by `allBodyLevel` to never override existing
structure.
- **Tag/One chunker fields (1.8):** `tag.go` sets `TopInt` from source
row index; `one.go` preserves `Positions`/`PDFPositions` from source
items. TSV multi-line RowNum fix: tracks `contentStart` for correct row
attribution.
- **Overlapped_percent normalization (2.6):**
`NormalizeOverlappedPercent` in `schema/chunker.go` mirrors Python
`common/float_utils.py:50-58` — accepts `[0,1)` fraction or `[0,90]`
percent, normalizes to canonical `[0,90]`.
- **Paragraph splitting (2.7):** aligned to Python flow `naive_merge` —
`CRLF` normalization, `splitKeepingDelimiter` preserves sentence
delimiters, single-section merge with token-budget-governed chunking.
- **chunk order:** sort by reading order
### Tokenizer
- **Phantom chunk filtering (Omission 2):** `isPhantomChunk` + filter
loop in `chunksFromTokenizerUpstream` skips zero-value ChunkDocs.
- **Batch size env var (Omission 3):** `embeddingBatchSize()` reads
`TOKENIZER_EMBEDDING_BATCH_SIZE`, defaults to 16.
- **Summary empty check (Diff 5):** `TrimSpace(s) != ""` → `s != ""`,
matching Python truthy check.
- **chunk_order_int all paths (Diff 8):** set unconditionally before
full_text/embedding branching.
- **Timeout default (Diff 10):** `600s` → `60s`, matching Python
`@timeout(60)`.
- **Small maxTokens truncation (Diff 14):** `truncateForEmbedding`
returns `""` when `maxTokens <= 10`, matching Python.
### Code review fixes
- Semaphore acquire moved before goroutine in `pdf_vision_dispatch.go`
(concurrency control)
- Context propagation in `pptx_tcadp.go` (cancellation support)
- Test resolver leak fix in `media_dispatch_test.go` (defer restore)
- Migration history comments removed per AGENTS.md
## Test plan
```
bash build.sh --test ./internal/parser/parser/... ./internal/ingestion/component/...
```
## Notes
- Migration diff tracking: `docs/migration_python_go_diff.md`
- Remaining gaps: Extractor component only (21 items)
2026-07-28 11:12:52 +08:00
|
|
|
// TestMaybeDispatchAudio_DefaultOutputFormatJson covers Parser 2.11:
|
|
|
|
|
// the default audio output_format must be "json" (matching Python
|
|
|
|
|
// parser.py:232 and AllowedOutputFormat["audio"]={"json"}).
|
|
|
|
|
func TestMaybeDispatchAudio_DefaultOutputFormatJson(t *testing.T) {
|
|
|
|
|
const want = "hello world"
|
|
|
|
|
drv := &audioTranscribeDriver{transcription: want}
|
|
|
|
|
orig := resolveTenantModelByType
|
|
|
|
|
defer func() { resolveTenantModelByType = orig }()
|
|
|
|
|
resolveTenantModelByType = func(ctx context.Context, db *gorm.DB, tenantID string, modelType entity.ModelType) (modelModule.ModelDriver, string, *modelModule.APIConfig, int, error) {
|
|
|
|
|
return drv, "asr-model", &modelModule.APIConfig{}, 0, nil
|
|
|
|
|
}
|
|
|
|
|
setups := defaultSetups()
|
|
|
|
|
// Do NOT set output_format — exercise the default path.
|
|
|
|
|
res, dispatched, err := maybeDispatchAudio(
|
|
|
|
|
context.Background(),
|
|
|
|
|
nil,
|
|
|
|
|
utility.FileTypeAURAL,
|
|
|
|
|
"test.mp3",
|
|
|
|
|
[]byte("fake-audio"),
|
|
|
|
|
map[string]any{"tenant_id": "t1"},
|
|
|
|
|
setups,
|
|
|
|
|
)
|
|
|
|
|
if err != nil {
|
|
|
|
|
t.Fatalf("maybeDispatchAudio: %v", err)
|
|
|
|
|
}
|
|
|
|
|
if !dispatched {
|
|
|
|
|
t.Fatal("expected dispatched=true for AURAL file")
|
|
|
|
|
}
|
|
|
|
|
if res.OutputFormat != "json" {
|
|
|
|
|
t.Fatalf("default OutputFormat = %q, want json", res.OutputFormat)
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
// TestMaybeDispatchMarkdownVision_EnhancesTables pins diff 2.5: markdown
|
|
|
|
|
// vision enhancement must also process items whose doc_type_kwd is "table"
|
|
|
|
|
// (Python checks {"image","table"} in parser/utils.py:181), not only "image".
|
|
|
|
|
// Before the fix the table item was skipped and never sent to the VLM.
|
|
|
|
|
func TestMaybeDispatchMarkdownVision_EnhancesTables(t *testing.T) {
|
|
|
|
|
origResolver := resolveTenantModelByType
|
|
|
|
|
defer func() { resolveTenantModelByType = origResolver }()
|
|
|
|
|
|
|
|
|
|
drv := &imagePromptCaptureDriver{}
|
2026-07-27 19:30:41 +08:00
|
|
|
resolveTenantModelByType = func(ctx context.Context, db *gorm.DB, tenantID string, modelType entity.ModelType) (modelModule.ModelDriver, string, *modelModule.APIConfig, int, error) {
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
return drv, "img-model", &modelModule.APIConfig{}, 0, nil
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
dispatched := parserDispatchResult{
|
|
|
|
|
OutputFormat: "json",
|
|
|
|
|
JSON: []map[string]any{
|
|
|
|
|
{"doc_type_kwd": "table", "image": "base64table", "text": ""},
|
|
|
|
|
},
|
|
|
|
|
}
|
|
|
|
|
|
2026-07-27 19:30:41 +08:00
|
|
|
ctx := t.Context()
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
res, handled, err := maybeDispatchMarkdownVision(
|
2026-07-27 19:30:41 +08:00
|
|
|
ctx,
|
|
|
|
|
dao.DB,
|
Fix(go): align ingestion pipeline with Python (parser/media dispatch + PDF coordinate chain + Chunker) (#17349)
## Summary
Aligns the Go ingestion pipeline with the Python implementation, closing
several behavioral gaps found during the Python→Go migration (tracked in
`docs/migration_python_go_diff.md`). Covers parser/media dispatch
alignment, the PDF coordinate-chain (preview images, outline→title,
chunk coordinate finalization), and the Chunker Token/QA batches below.
Commits are grouped as follows.
### 1. Fix parser params (c524f450e)
Fixes parser/media wiring and several dispatch gaps:
- **docx/pdf vision dispatch**: correct parameter handling and VLM
invocation.
- **markdown vision (diff 2.5)**: also enhance items whose
`doc_type_kwd` is `table`, not only `image` (parser/utils.py:181).
- **media audio (diff 2.11)**: when `output_format` is `json`, carry the
ASR transcription as a JSON item instead of only the `Text` field (the
Invoke switch had no `json` branch and dropped it).
- **email (diff 2.2)**: default `output_format` is `json`
(parser.py:212), not `text`.
- **tokenizer**: handle empty/whitespace-only names; trim before
embedding.
- **extractor**: tag-matching parameter wiring.
- **split**: keyword-split regex now covers CJK/English separators.
- **parser.go**: parser-param plumbing.
### 2. fix parser gap (373537da1)
Image dispatch now mirrors `rag/app/picture.py:chunk()`:
- Always OCR the image (PaddleOCR or local ONNX).
- When OCR text is short, also call VLM (`describe`) and combine `OCR +
VLM` text.
- Emits a **structured JSON item** carrying the image data-URI and
`doc_type_kwd:"image"`, instead of a bare `Text` string. This fixes the
payload being rejected downstream by OneChunker/TokenChunker (JSON=nil).
### 3. PDF coordinate-chain fixes (55367a820, 727f8167c)
Closes three items from the migration tracker in the
chunker/tokenizer/task layer:
- **(Chunker-1.3) `restore_pdf_text_previews`** — `needsCrop` now also
returns true for `text` chunks that carry PDF positions
(`pdfcrop_cgo.go`), so text blocks get a rendered preview image uploaded
to storage via `imageUploadDecorator`/`ChunkImageUploader`, matching
Python `restore_pdf_text_previews` + `image2id`.
- **(Chunker-1.5) PDF outline → title levels** — `title.go` adds
`outlineSimilarity` (rune-bigram Jaccard, mirroring
`common.py:_outline_similarity`), `resolveOutlineLevels` (matches text
lines to outline entries at similarity > 0.8, with a sparse guard
`len(outline)/len(records) <= 0.03`), and `outlineFromInputs` (reads
`file.outline`). Wired into `newLevelContext` in both `group.go` and
`hierarchy.go`; falls back to the title-shape heuristic when no outline
is present.
- **(Tokenizer-(T)1) `finalize_pdf_chunk`** — the coordinate →
`position_int`/`page_num_int`/`top_int` conversion is owned by the task
layer (`processChunkPositions`→`AddPositions`), which runs *after* the
tokenizer and consumes the tokenizer-owned fields. The tokenizer only
preserves the raw `positions`/`_pdf_positions` (no duplicate
conversion), pinned by `TestChunkDocsToMaps_PreservesPDFPositions`.
### 4. Integration test made environment-free
(`internal/ingestion/task/pipeline_real_integration_test.go`)
- Removed the `//go:build integration` tag so the contract tests run
under the default `build.sh --test` (which does not pass `-tags
integration`).
- External dependencies replaced with in-memory substitutes so no
MySQL/MinIO/ES is required:
- MySQL → on-disk sqlite (`glebarez/sqlite`) with the needed tables
auto-migrated.
- MinIO → `storage.NewMemoryStorage()`.
- Elasticsearch → chunks captured via `WithInsertFunc` instead of
`engine.InsertChunks`/`Search`.
- `requireTokenizerPool` still skips gracefully when the native
tokenizer pool is unavailable; `WithLogCreateFunc(noop)` avoids
depending on the operation-log table.
- Added `taskChunkFieldEqualsStr` to tolerate `kb_id` being a
`[]string`/`[]any` in the raw chunk payload (the search engine flattens
it to a string on read).
### 5. TokenChunker alignment — Batch 1
(`internal/ingestion/component/chunker/token.go`)
Closes four Chunker items from the migration tracker:
- **(Chunker-2.1) sentence delimiter** — the boundary regex now also
breaks on ASCII `!`/`?`. Extracted to a package-level `var
sentenceDelimiter` and used in `mergeByTokenSize`, matching Python's
full delimiter set.
- **(Chunker-2.2) overlap tag leakage** — when a new chunk starts, its
overlap prefix is taken from the previous chunk *after* `removeTag`, in
both the text path (`mergeByTokenSize`) and the JSON path
(`mergeByTokenSizeFromJSON`). Parser tags (`@@…##`) no longer leak into
the overlap region (mirrors `nlp/__init__.py:1181`).
- **(Chunker-2.11) empty-text merge** — merging a non-empty chunk into
an empty previous chunk now assigns the text directly instead of being
skipped (`mergeByTokenSizeFromJSON`), mirroring
`token_chunker.py:236-239`.
- **(Chunker-2.4) overlap token counting** —
`takeFromEnd`/`takeFromStart` now count tokens exactly via `tokenizeStr`
instead of the 4-bytes/token heuristic, fixing over-counting for CJK
text.
### 6. QA Chunker alignment — Batch 2
(`internal/ingestion/component/chunker/qa.go` + `schema`)
Closes three Chunker items from the migration tracker:
- **(Chunker-2.13) default language** — an empty `lang` now defaults to
Chinese prefixes (`问题:`/`回答:`) instead of English, matching `qa.py:299`.
- **(Chunker-2.12) `rmQAPrefix` regex** — the separator is changed to
`[\t:: ]+` (one-or-more), matching `qa.py:241`, so multiple separators
(e.g. `Q:: answer`) are fully stripped.
- **(Chunker-1.8 QA) missing chunk fields** — QA chunks now preserve:
- `top_int` — the source row/record index, threaded through the
tab/csv/markdown extractors (mirrors `qa.py` `beAdoc(..., row_num=i)`);
- `image` + `doc_type_kwd:"image"`;
- `_pdf_positions` / `positions` carried from the upstream JSON item.
`schema.ChunkDoc` gains a `TopInt []int` field (serialized as `top_int`,
registered in `UnmarshalJSON`). Note: the Tag/Table/Presentation/One
chunker field gaps under 1.8 remain pending.
## Test plan
- Added/updated unit tests: `pdfcrop_cgo_test.go` (`TestNeedsCrop`,
`TestRestorePDFTextPreview`), `title_test.go`
(`TestResolveOutlineLevels`, `TestResolveOutlineLevels_SparseGuard`,
`TestNewLevelContext_OutlineBranch`, `TestOutlineFromInputs`),
`tokenizer_unit_test.go` (`TestChunkDocsToMaps_PreservesPDFPositions`),
`token_pdfpos_test.go`.
- **Batch 1** — `token_batch1_test.go`:
`TestSentenceDelimiterMatchesBangAndQuestion`,
`TestMergeByTokenSizeFromJSON_OverlapStripsTags`,
`TestMergeByTokenSizeFromJSON_EmptyPrevKeepsChunk`,
`TestTakeFromEndRespectsTokenCount`,
`TestTakeFromStartRespectsTokenCount`.
- **Batch 2** — `qa_batch2_test.go`:
`TestQAChunker_DefaultLangIsChinese`,
`TestRmQAPrefixStripsMultipleSeparators`, `TestQAChunker_SetsTopInt`,
`TestQAChunker_CarriesImageAndPositions`. Existing `qa_test.go`
expectations were updated to the corrected language default / separator
behavior.
- `pipeline_real_integration_test.go`
(`TestPipelineExecutor_Run_RealCanvasDSL_UsesGeneralPipeline`,
`TestPipelineExecutor_Run_RealPDF_ProducesIndexedChunks`,
`TestRunPipeline_RealPipelineOutput_ProducesIndexFields`) now runs
without any external service.
- `bash build.sh --test ./internal/ingestion/...` passes.
- No files deleted.
2026-07-24 21:06:38 +08:00
|
|
|
utility.FileTypeMarkdown,
|
|
|
|
|
dispatched,
|
|
|
|
|
map[string]any{"tenant_id": "t1"},
|
|
|
|
|
)
|
|
|
|
|
if err != nil {
|
|
|
|
|
t.Fatalf("maybeDispatchMarkdownVision: %v", err)
|
|
|
|
|
}
|
|
|
|
|
if !handled {
|
|
|
|
|
t.Fatalf("expected handled=true for markdown with a table image")
|
|
|
|
|
}
|
|
|
|
|
if len(res.JSON) != 1 {
|
|
|
|
|
t.Fatalf("JSON len = %d, want 1", len(res.JSON))
|
|
|
|
|
}
|
|
|
|
|
// The table item must have been sent to the VLM and its description appended.
|
|
|
|
|
if got, _ := res.JSON[0]["text"].(string); got != "captured" {
|
|
|
|
|
t.Fatalf("table item text = %q, want %q (table items must be vision-enhanced)", got, "captured")
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// TestDefaultEmailOutputFormatIsJSON pins diff 2.2: the email family default
|
|
|
|
|
// output_format must be "json" (matching Python parser.py:212), not "text".
|
|
|
|
|
// With "text" the structured email fields (from/to/subject/attachments/...) are
|
|
|
|
|
// flattened into a blob and lost downstream.
|
|
|
|
|
func TestDefaultEmailOutputFormatIsJSON(t *testing.T) {
|
|
|
|
|
got, _ := defaultSetups()["email"]["output_format"].(string)
|
|
|
|
|
if got != "json" {
|
|
|
|
|
t.Fatalf("email default output_format = %q, want json", got)
|
|
|
|
|
}
|
|
|
|
|
}
|