Update multi_column.py

OlgaDumina · web-flow · commit 006c611227c4 · 2025-03-04T11:53:04.000+01:00
It solves an issue with wrong column extraction, while the lines in the column weren't joined properly. The test set in line 246 was created from the old value of prects[0], but compared with new prect0.
diff --git a/pymupdf4llm/pymupdf4llm/helpers/multi_column.py b/pymupdf4llm/pymupdf4llm/helpers/multi_column.py
@@ -248,6 +248,7 @@ def join_rects_phase3(bboxes, path_rects):
                     )
                     if test == set((tuple(prect0), tuple(prect1))):
                         prect0 |= prect1
+                        prects[0] = prect0
                         del prects[i]
                         repeat = True
             new_rects.append(prect0)

Original file line number	Diff line number	Diff line change
`@@ -248,6 +248,7 @@ def join_rects_phase3(bboxes, path_rects):`
`248`	`248`	`)`
`249`	`249`	`if test == set((tuple(prect0), tuple(prect1))):`
`250`	`250`	`prect0 \|= prect1`
	`251`	`+ prects[0] = prect0`
`251`	`252`	`del prects[i]`
`252`	`253`	`repeat = True`
`253`	`254`	`new_rects.append(prect0)`