Files
ragflow/internal/deepdoc/parser/pdf/table_rotate_integration_test.go
Jack d12fd3b79d feat: parser pages range and parse type validation for dataset/document (#17293)
## Summary

Adds page-range parsing support to the Go-native pipeline path and
introduces strict `parse_type` validation for both dataset and document
update endpoints.

## What changed

### Pages range parsing
- **`internal/utility/pdf_pages.go`** — `NormalizePDFPages`: normalizes
raw page ranges (list of `[from,to]` 1-indexed inclusive ranges) into
sorted, merged, deduplicated `[][]int`. Invalid ranges are dropped.
- **`internal/ingestion/pipeline/pdf_pages.go`** —
`NormalizeParserConfigPages`: walks any parser_config map and normalizes
`"pages"` values under every component → filetype setup, so the
persisted config always carries clean, merged ranges.
- **`internal/deepdoc/parser/pdf/parser.go`** — integrates
`resolvePagesToProcess` to filter parsed PDF pages by the configured
ranges.
- Pipeline integration (parser pages):
`internal/parser/parser/pdf_parser_common.go`, `chunk_process.go`, plus
associated e2e and unit tests.

### Parse type validation (shared logic)
- **`internal/service/parser_mode.go`** (new) — `ValidateParseTypeMode`:
shared function that validates `parse_type` (1=BuiltIn/parser_id,
2=Pipeline/pipeline_id) and ensures the corresponding field is present.
Used by both dataset and document update endpoints.
- **`internal/service/dataset/crud.go`** / `update.go` — replaces inline
`isPipelineMode`/`isBuiltinMode` computation with the shared
`service.ValidateParseTypeMode`.
- **`internal/service/document/document_dataset_update.go`** — adds
strict `parse_type` validation in `validateDatasetDocumentUpdate`,
simplifies the reparse logic to a two-way switch (isBuiltin/isPipeline)
now that parse_type is always valid.
- **`internal/service/document/document.go`** — adds `ParseType` field
to `UpdateDatasetDocumentRequest`.
- **`internal/service/document/document_dataset_update.go`** —
`updateDocumentParserConfig` fallback path when DSL loading fails.
- **`internal/service/parser_mode_test.go`** (new) — test coverage for
nil, invalid, and missing-field scenarios.

### Frontend
- **`web/src/interfaces/request/document.ts`** — adds `parseType` to
`IChangeParserRequestBody`.
- **`web/src/hooks/use-document-request.ts`** —
`useSetDocumentPipelineParser` sends `parse_type` in the PATCH payload.
- **`web/src/pages/dataset/dataset/use-change-document-parser.ts`** —
Go/Python branching for the document parser config dialog.
-
**`web/src/components/document-pipeline-dialog/use-document-pipeline-form.ts`**
— `buildSubmitData` returns `parseType` (bugfix: was dropped from the
return value).

### Test changes
- **Removed**: 2 tests that verified the old "mutually exclusive" error
(replaced by `ValidateParseTypeMode` coverage).
- **Modified**: 6 tests across document and dataset packages to include
`ParseType` in request structs.
- **Added**: new e2e tests for pages parsing (`pages_e2e_test.go`,
`pdf_parser_pages_e2e_test.go`) and unit tests for `NormalizePDFPages`,
`NormalizeParserConfigPages`, `resolvePagesToProcess`.

## Backward compatibility
- The `parse_type` field is **required** when `parser_id` or
`pipeline_id` is sent. This changes the contract for both dataset and
document PATCH endpoints, but aligns the Go backend with the existing
frontend behavior (the frontend already sends `parse_type`). Callers
that omit `parse_type` when updating parser/pipeline selections will
receive a clear error message.
- Existing callers that only update fields like `name`, `enabled`, or
`meta_fields` are unaffected.
- Test updates ensure all known call sites are compliant.
2026-07-23 19:57:27 +08:00

201 lines
5.0 KiB
Go

//go:build cgo && manual
package pdf
import (
"context"
"os"
"path/filepath"
"ragflow/internal/common"
inf "ragflow/internal/deepdoc/parser/pdf/inference"
tbl "ragflow/internal/deepdoc/parser/pdf/table"
pdf "ragflow/internal/deepdoc/parser/pdf/type"
util "ragflow/internal/deepdoc/parser/pdf/util"
"testing"
)
// TestTableRotation_Integration validates rotation detection with real DeepDoc.
//
// Prerequisites:
// - DeepDoc running at localhost:9390 (or set DEEPDOC_URL)
// - Test PDF: testdata/pdfs/table_rotation_test.pdf (generated by tools/generate_rotated_table_pdf.py)
//
// Run:
//
// CGO_CFLAGS="..." CGO_LDFLAGS="..." \
// go test -tags 'cgo,manual' -run TestTableRotation_Integration -v -count=1
func TestTableRotation_Integration(t *testing.T) {
pdfPath := filepath.Join("testdata", "pdfs", "table_rotation_test.pdf")
if _, err := os.Stat(pdfPath); os.IsNotExist(err) {
t.Skipf("test PDF not found: %s (run tools/generate_rotated_table_pdf.py first)", pdfPath)
}
baseURL := common.GetEnv(common.EnvDeepDocURL)
if baseURL == "" {
baseURL = "http://localhost:9390"
}
dd, err := inf.NewClient(baseURL)
if err != nil {
t.Fatal(err)
}
if !dd.Health() {
t.Fatalf("DeepDoc not available at %s", baseURL)
}
t.Logf("DeepDoc available at %s", baseURL)
// Open PDF
data, err := os.ReadFile(pdfPath)
if err != nil {
t.Fatal(err)
}
eng, err := NewEngine(data)
if err != nil {
t.Fatal(err)
}
defer eng.Close()
pageCount, _ := eng.PageCount()
t.Logf("PDF: %d pages", pageCount)
cfg := pdf.DefaultParserConfig()
autoRotate := true
cfg.AutoRotateTables = &autoRotate
_ = NewParser(cfg) // verify construction does not panic
for pg := 0; pg < pageCount; pg++ {
pageImg, err := RenderPageToImage(eng, pg)
if err != nil {
t.Fatalf("render page %d: %v", pg, err)
}
regions, err := dd.DLA(context.Background(), pageImg)
if err != nil {
t.Fatalf("DLA page %d: %v", pg, err)
}
tableCount := 0
for _, r := range regions {
if r.Label != "table" {
continue
}
tableCount++
// Crop table region
cropped, err := util.CropImageRegion(pageImg, r)
if err != nil {
t.Errorf(" crop table %d: %v", tableCount, err)
continue
}
// Evaluate rotation
angle, _, scores := tbl.EvaluateTableOrientation(context.Background(), cropped, dd)
t.Logf(" Page %d Table %d: %dx%d, bestAngle=%d°, scores: 0=%.3f 90=%.3f 180=%.3f 270=%.3f",
pg, tableCount, cropped.Bounds().Dx(), cropped.Bounds().Dy(),
angle,
scores[0], scores[90], scores[180], scores[270])
// Verify: page 0 should be ~0°, page 1 should be ~90°
if pg == 0 && angle != 0 {
t.Errorf("Page 0 normal table: expected 0°, got %d°", angle)
}
// Page 1 has the rotated table - expect 90° (or 270° depending on DLA bbox)
if pg == 1 {
t.Logf(" NOTE: Page 1 rotated table detected as %d° (expect 90 or 270)", angle)
// Verify TSR returns labels (6th element in bbox array).
testCells, tsrErr := dd.TSR(context.Background(), cropped)
if tsrErr == nil && len(testCells) > 0 {
hasLabel := false
for _, c := range testCells {
if c.Label != "" {
hasLabel = true
break
}
}
if !hasLabel {
t.Error("TSR returned cells without labels")
} else {
t.Logf(" TSR labels OK: %d cells", len(testCells))
}
}
}
}
t.Logf("Page %d: %d tables detected", pg, tableCount)
}
}
// TestTableRotation_Stability runs rotation detection on a sample real PDF
// and verifies the pipeline doesn't crash. Set BATCH_COUNT to limit.
func TestTableRotation_Stability(t *testing.T) {
baseURL := common.GetEnv(common.EnvDeepDocURL)
if baseURL == "" {
baseURL = "http://localhost:9390"
}
dd, err := inf.NewClient(baseURL)
if err != nil {
t.Fatal(err)
}
if !dd.Health() {
t.Fatalf("DeepDoc not available at %s", baseURL)
}
realDir := filepath.Join("testdata", "real_pdfs")
entries, err := os.ReadDir(realDir)
if err != nil {
t.Skipf("no real PDFs: %v", err)
}
count := 0
maxCount := 3 // sample size
for _, e := range entries {
if e.IsDir() || filepath.Ext(e.Name()) != ".pdf" {
continue
}
if count >= maxCount {
break
}
data, err := os.ReadFile(filepath.Join(realDir, e.Name()))
if err != nil {
continue
}
eng, err := NewEngine(data)
if err != nil {
continue
}
pageImg, err := RenderPageToImage(eng, 0)
eng.Close()
if err != nil {
continue
}
regions, _ := dd.DLA(context.Background(), pageImg)
tables := 0
rotated := 0
for _, r := range regions {
if r.Label != "table" {
continue
}
tables++
cropped, err := util.CropImageRegion(pageImg, r)
if err != nil {
t.Errorf(" %s crop table: %v", e.Name(), err)
continue
}
if cropped == nil {
continue
}
angle, _, _ := tbl.EvaluateTableOrientation(context.Background(), cropped, dd)
if angle != 0 {
rotated++
t.Logf(" %s: rotated table detected (angle=%d°)", e.Name(), angle)
}
}
t.Logf(" %s: %d tables, %d rotated", e.Name(), tables, rotated)
count++
}
t.Logf("Sampled %d real PDFs", count)
}