Upload 2 files

v-corpus-zh/Navel/近月少女的礼仪/近月少女的礼仪/luna.txt.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11d6388b39eff759323408005d4013f39a40bef9971c3bcadc1f5391e6657f18
+size 602295

v-corpus-zh/Navel/近月少女的礼仪/近月少女的礼仪/tmp_luna.py

@@ -0,0 +1,111 @@
+import fitz
+
+print(fitz.__doc__)
+
+a = r'E:\tmp\jysn\luna_v1.0s3.pdf'
+
+_end = {'」', '。', '…', '？', '！'}
+
+_lines = []
+
+doc = fitz.open(a)  # open a document
+TIT = False
+for page in doc:  # iterate the document pages
+    text = page.get_text("blocks")  # get plain text encoded as UTF-8
+    if len(text) <= 0:
+        continue
+    tmp = text[0][4]
+    tmp = tmp.split(' ')
+    if len(tmp) != 2 or not tmp[0].isdigit() or tmp[1] != '\n':
+        continue
+    # print(tmp)
+    if tmp[0] == '732':
+        break
+    i = 0
+    _l = len(text) - 2
+    while i <= _l:
+        i += 1  # 跳过第一行的页码
+        if TIT:
+            TIT = False
+            line = '章名：' + text[i][4].strip()
+            if line[-1] not in _end:
+                line += '。'
+            print('跨页TIT', line)
+        if text[i][4].startswith('TIT \n'):  # 小标题
+            line = text[i][4][5:].strip()
+            if line:
+                line = '章名：' + line
+            else:
+                TIT = True
+                continue  # 服了，这居中标记还能跨页。。。
+            if line[-1] not in _end:
+                line += '。'
+            print(line)
+        else:
+            line = text[i][4].strip()
+        if not line:
+            continue
+        if line.startswith('BGT') and line.endswith(_lines[-1]):  # 章节之间的空白
+            _lines[-1] = '卷名：' + _lines[-1] + '。'
+            continue
+        _lines.append(line)
+
+data = []
+
+左括号 = False
+cache = ''
+
+i = -1
+_l = len(_lines) - 2
+while i <= _l:
+    i += 1
+    line = _lines[i]
+    # print(line)
+    # tmp = line[-1]
+    # _end[tmp] = _end.get(tmp, 0) + 1
+    line = line.replace('：\n「', '：「')
+    line = line.replace('\n', '\\n')
+    if '：「' in line:
+        if cache != '':
+            print('误判', (cache, line))  # 之前误判的
+            if 左括号:
+                data.append(cache)
+            else:
+                data.append('旁白：' + cache)
+            cache = ''
+
+        左括号 = True
+
+    if not 左括号 and '「' in line:
+        print('待处理', repr(line))
+        if line.count('「') != line.count('」'):
+            if '：' not in line and ':' not in line:
+                pass
+            else:
+                print('异常中断')
+                break  # 手动处理错误
+
+        左括号 = False
+
+    if not 左括号:
+        if line[-1] in _end:
+            data.append('旁白：' + cache + line)
+            cache = ''
+        else:
+            cache += line
+    else:
+        if line[-1] == '」':
+            data.append(cache + line)
+            cache = ''
+            左括号 = False
+        else:
+            cache += line
+
+
+# sorted_word_freq = sorted(_end.items(), key=lambda x: x[1], reverse=True)
+
+b = r'D:\datasets\tmp'
+k = 'luna'
+v = data
+with open(b + f'\\{k}.txt', 'w', encoding='utf-8') as f:
+    f.write('\n'.join(v))