Update app.py

app.py

@@ -1,3 +1,4 @@
+import difflib
 import re
 
 HYPHENS = set(['\u002D','\uFF0D','\u2010','\u2011','\u2013','\u2014'])
@@ -98,22 +99,109 @@ def build_active_text_and_map(original, flags):
     index_map = [i for i, _ in active_chars]
     return active_text, index_map
 
-src25 = "abc(def)123アイウ〔ghi〕fi⼀zero-test \nabc\n(def)123アイウ〔ghi〕fi⼀zero-test \nabc"
-dst25 = "xyz(www)789カキクケコ〔jkl〕fl⼆one+test \nxyz\n(www)789カキクケコ〔jkl〕fl⼆one+test \nxyz"
-settings25 = {
-    'ignore_alpha': True, 'ignore_digit': True, 'ignore_kana': True, 'ignore_yakumono': True,
-    'ignore_ligature': True, 'ignore_kangxi': True, 'ignore_brackets_with_content': True,
-    'ignore_kanjinum': True, 'ignore_hyphen': True, 'ignore_kumimoji': True,
-    'ignore_space': True, 'ignore_newline': True
-}
-
-src_flags = build_active_flags(src25, settings25)
-dst_flags = build_active_flags(dst25, settings25)
-src_active, _ = build_active_text_and_map(src25, src_flags)
-dst_active, _ = build_active_text_and_map(dst25, dst_flags)
-print(f"src_active: '{src_active}'")
-print(f"dst_active: '{dst_active}'")
-src_removed = ''.join(ch for i, ch in enumerate(src25) if not src_flags[i])
-dst_removed = ''.join(ch for i, ch in enumerate(dst25) if not dst_flags[i])
-print(f"src_removed: '{src_removed}'")
-print(f"dst_removed: '{dst_removed}'")
+def build_removed_text_and_map(original, flags):
+    removed_chars = [(i, ch) for i, ch in enumerate(original) if not flags[i]]
+    removed_text = ''.join(ch for _, ch in removed_chars)
+    index_map = [i for i, _ in removed_chars]
+    return removed_text, index_map
+
+def merge_ranges(ranges):
+    merged = []
+    for s, e in sorted(ranges):
+        if s >= e: continue
+        if merged and s <= merged[-1][1]:
+            merged[-1] = (merged[-1][0], max(merged[-1][1], e))
+        else:
+            merged.append([s, e])
+    return [(s, e) for s, e in merged]
+
+def compare(src, dst, settings):
+    src_flags = build_active_flags(src, settings)
+    dst_flags = build_active_flags(dst, settings)
+
+    src_active, src_map = build_active_text_and_map(src, src_flags)
+    dst_active, dst_map = build_active_text_and_map(dst, dst_flags)
+
+    src_diff_ranges = []
+    dst_diff_ranges = []
+
+    # 有効文字の比較（両方空なら差分なし）
+    if src_active != '' or dst_active != '':
+        matcher = difflib.SequenceMatcher(None, src_active, dst_active, autojunk=False)
+        for tag, i1, i2, j1, j2 in matcher.get_opcodes():
+            if tag in ('replace', 'delete') and i2 > i1:
+                src_diff_ranges.append((src_map[i1], src_map[i2-1] + 1))
+            if tag in ('replace', 'insert') and j2 > j1:
+                dst_diff_ranges.append((dst_map[j1], dst_map[j2-1] + 1))
+
+    # 除去文字の比較
+    src_removed, src_removed_map = build_removed_text_and_map(src, src_flags)
+    dst_removed, dst_removed_map = build_removed_text_and_map(dst, dst_flags)
+
+    if src_removed != dst_removed and (src_active != '' or dst_active != ''):
+        # 除去文字同士をdifflibで比較
+        rm_matcher = difflib.SequenceMatcher(None, src_removed, dst_removed, autojunk=False)
+        for tag, i1, i2, j1, j2 in rm_matcher.get_opcodes():
+            if tag in ('replace', 'delete') and i2 > i1:
+                src_diff_ranges.append((src_removed_map[i1], src_removed_map[i2-1] + 1))
+            if tag in ('replace', 'insert') and j2 > j1:
+                dst_diff_ranges.append((dst_removed_map[j1], dst_removed_map[j2-1] + 1))
+
+    return merge_ranges(src_diff_ranges), merge_ranges(dst_diff_ranges)
+
+def apply_highlight(text, ranges):
+    result = []
+    prev = 0
+    for s, e in ranges:
+        result.append(text[prev:s])
+        result.append('"' + text[s:e] + '"')
+        prev = e
+    result.append(text[prev:])
+    return ''.join(result)
+
+test_cases = [
+    ("TEST-01", "吾輩は猫である。名前はまだない。", "吾輩は猫である。名前はまだない。", {}),
+    ("TEST-02", "吾輩は猫である。", "吾輩は犬である。", {}),
+    ("TEST-03", "abcdef", "abcXYZdef", {}),
+    ("TEST-04", "abcXYZdef", "abcdef", {}),
+    ("TEST-08", "吾輩は、猫である。「名前」は（まだ）ない！", "吾輩は猫である名前はまだない",
+     {'ignore_yakumono': True}),
+    ("TEST-09", "(abc)テスト", "abcテスト",
+     {'ignore_yakumono': True}),
+    ("TEST-19", "abc-def－ghi‐jkl‑mno–pqr—stu", "abcdefghijklmnopqrstu",
+     {'ignore_hyphen': True}),
+    ("TEST-21", "abc def　ghi", "abcdefghi",
+     {'ignore_space': True}),
+    ("TEST-22", "abc\ndef\nghi", "abcdefghi",
+     {'ignore_newline': True}),
+    ("TEST-23", "吾輩は(猫)である。名前は【まだ】ない！", "吾輩はである。名前はない！",
+     {'ignore_yakumono': True, 'ignore_brackets_with_content': True}),
+    ("TEST-25",
+     "abc(def)123アイウ〔ghi〕fi⼀zero-test \nabc\n(def)123アイウ〔ghi〕fi⼀zero-test \nabc",
+     "xyz(www)789カキクケコ〔jkl〕fl⼆one+test \nxyz\n(www)789カキクケコ〔jkl〕fl⼆one+test \nxyz",
+     {'ignore_alpha': True, 'ignore_digit': True, 'ignore_kana': True, 'ignore_yakumono': True,
+      'ignore_ligature': True, 'ignore_kangxi': True, 'ignore_brackets_with_content': True,
+      'ignore_kanjinum': True, 'ignore_hyphen': True, 'ignore_kumimoji': True,
+      'ignore_space': True, 'ignore_newline': True}),
+    ("TEST-30", "abc def", "abcdef", {}),
+    ("TEST-35", "abcXYZdef", "abcdef", {}),
+    ("TEST-37", "abc-def－ghi‐jkl‑mno–pqr—stu", "abcdefghijklmnopqrstu",
+     {'ignore_hyphen': True}),
+    # 比較元と比較先を入れ替えたケース
+    ("TEST-08-Rev", "吾輩は猫である名前はまだない", "吾輩は、猫である。「名前」は（まだ）ない！",
+     {'ignore_yakumono': True}),
+    ("TEST-19-Rev", "abcdefghijklmnopqrstu", "abc-def－ghi‐jkl‑mno–pqr—stu",
+     {'ignore_hyphen': True}),
+    ("TEST-21-Rev", "abcdefghi", "abc def　ghi",
+     {'ignore_space': True}),
+    ("TEST-22-Rev", "abcdefghi", "abc\ndef\nghi",
+     {'ignore_newline': True}),
+    ("TEST-23-Rev", "吾輩はである。名前はない！", "吾輩は(猫)である。名前は【まだ】ない！",
+     {'ignore_yakumono': True, 'ignore_brackets_with_content': True}),
+]
+
+for name, src, dst, settings in test_cases:
+    src_ranges, dst_ranges = compare(src, dst, settings)
+    print(f"\n=== {name} ===")
+    print(f"  比較元: {apply_highlight(src, src_ranges)}")
+    print(f"  比較先: {apply_highlight(dst, dst_ranges)}")