Files
ragflow/internal/ingestion/component/docx_vision_dispatch_test.go
Jack 554925b583 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

182 lines
6.3 KiB
Go

//
// 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"
"sync"
"testing"
"ragflow/internal/entity"
modelModule "ragflow/internal/entity/models"
"ragflow/internal/utility"
)
// docxVisionFakeDriver satisfies modelModule.ModelDriver but never reaches the
// network: docxVisionCaptureInvoker intercepts the call before the driver.
type docxVisionFakeDriver struct {
modelModule.ModelDriver
}
// docxVisionCaptureInvoker records the requested image and returns a fixed
// description, mirroring the markdown-vision test's capture driver.
type docxVisionCaptureInvoker struct {
mu sync.Mutex
images []string
captured []modelModule.Message
}
func (c *docxVisionCaptureInvoker) invoke(
ctx context.Context,
driver modelModule.ModelDriver,
modelName string,
messages []modelModule.Message,
apiConfig *modelModule.APIConfig,
) (*modelModule.ChatResponse, error) {
c.mu.Lock()
c.captured = append(c.captured, messages...)
// Pull the data URI out of the second content part.
if parts, ok := messages[0].Content.([]interface{}); ok && len(parts) >= 2 {
if img, ok := parts[1].(map[string]any); ok {
if url, ok := img["image_url"].(map[string]any); ok {
if u, ok := url["url"].(string); ok {
c.images = append(c.images, u)
}
}
}
}
c.mu.Unlock()
ans := "a diagram of a pipeline"
return &modelModule.ChatResponse{Answer: &ans}, nil
}
// TestMaybeDispatchDOCXVision_EnhancesJSONImages verifies Diff 2.4: DOCX vision
// enhancement must trigger on the JSON output path (like Python's
// enhance_media_sections_with_vision in parser.py:_doc) and must NOT trigger on
// the markdown path. Image items with a non-empty `image` field get their VLM
// description appended to `text`; table items (no image) and text items are
// left untouched.
func TestMaybeDispatchDOCXVision_EnhancesJSONImages(t *testing.T) {
origResolver := resolveTenantModelByType
origInvoker := visionChatInvoker
origPrompt := docxVisionPromptBuilder
defer func() {
resolveTenantModelByType = origResolver
visionChatInvoker = origInvoker
docxVisionPromptBuilder = origPrompt
}()
resolveTenantModelByType = func(tenantID string, modelType entity.ModelType) (modelModule.ModelDriver, string, *modelModule.APIConfig, int, error) {
return &docxVisionFakeDriver{}, "docx-vision-model", &modelModule.APIConfig{}, 0, nil
}
invoker := &docxVisionCaptureInvoker{}
visionChatInvoker = invoker.invoke
docxVisionPromptBuilder = func(string, string) (string, error) { return "describe the figure", nil }
dispatched := parserDispatchResult{
OutputFormat: "json",
DocType: "docx",
JSON: []map[string]any{
{"text": "Intro paragraph", "image": nil, "doc_type_kwd": "text"},
{"text": "", "image": "aGVsbG8taW1hZ2U=", "doc_type_kwd": "image"},
{"text": "<table></table>", "image": nil, "doc_type_kwd": "table"},
},
}
res, handled, err := maybeDispatchDOCXVision(
context.Background(),
utility.FileTypeDOCX,
dispatched,
map[string]any{"tenant_id": "t1"},
defaultSetups(),
)
if err != nil {
t.Fatalf("maybeDispatchDOCXVision: unexpected error: %v", err)
}
if !handled {
t.Fatal("handled = false, want true (JSON image item should be enhanced)")
}
if len(res.JSON) != 3 {
t.Fatalf("JSON len = %d, want 3", len(res.JSON))
}
if got := res.JSON[0]["text"].(string); got != "Intro paragraph" {
t.Errorf("text item text = %q, want unchanged", got)
}
// image item: VLM description appended to (empty) text.
if got, _ := res.JSON[1]["text"].(string); got != "a diagram of a pipeline" {
t.Errorf("image item text = %q, want appended VLM description", got)
}
if got, _ := res.JSON[2]["text"].(string); got != "<table></table>" {
t.Errorf("table item text = %q, want unchanged (no image)", got)
}
if len(invoker.images) != 1 {
t.Fatalf("vision invoker called %d times, want 1 (only the image item)", len(invoker.images))
}
if want := "data:image/png;base64,aGVsbG8taW1hZ2U="; invoker.images[0] != want {
t.Errorf("vision image data URI = %q, want %q", invoker.images[0], want)
}
}
// TestMaybeDispatchDOCXVision_JSONOnly verifies Diff 2.4: the markdown output
// path must NOT be enhanced (Python's markdown/docx branch performs no vision
// enrichment). A markdown result with embedded figures is returned untouched.
func TestMaybeDispatchDOCXVision_JSONOnly(t *testing.T) {
origResolver := resolveTenantModelByType
origInvoker := visionChatInvoker
defer func() {
resolveTenantModelByType = origResolver
visionChatInvoker = origInvoker
}()
called := false
resolveTenantModelByType = func(tenantID string, modelType entity.ModelType) (modelModule.ModelDriver, string, *modelModule.APIConfig, int, error) {
called = true
return &docxVisionFakeDriver{}, "m", &modelModule.APIConfig{}, 0, nil
}
visionChatInvoker = func(ctx context.Context, d modelModule.ModelDriver, m string, msgs []modelModule.Message, c *modelModule.APIConfig) (*modelModule.ChatResponse, error) {
called = true
ans := "x"
return &modelModule.ChatResponse{Answer: &ans}, nil
}
dispatched := parserDispatchResult{
OutputFormat: "markdown",
DocType: "docx",
Markdown: "![Image](data:image/png;base64,abc)",
File: map[string]any{"figures": []map[string]any{{"image": "abc", "marker": "x"}}},
}
res, handled, err := maybeDispatchDOCXVision(
context.Background(),
utility.FileTypeDOCX,
dispatched,
map[string]any{"tenant_id": "t1"},
defaultSetups(),
)
if err != nil {
t.Fatalf("maybeDispatchDOCXVision: unexpected error: %v", err)
}
if handled {
t.Error("handled = true, want false (markdown path must not be enhanced)")
}
if called {
t.Error("vision model was resolved/invoked on the markdown path")
}
if res.Markdown != "![Image](data:image/png;base64,abc)" {
t.Errorf("markdown mutated: %q", res.Markdown)
}
}