Create write_markup.py

write_markup.py

@@ -0,0 +1,128 @@
+import os
+import json
+import fitz
+from glob import glob
+
+# 本脚本默认放在笔记数据目录中 保存的PDF文件在output文件夹里
+note_path = './notebooks'
+dst_pdf_path = './output'
+
+
+def path_join(*paths):
+    path = os.path.join(*paths)
+    path = os.path.normcase(path)
+    return path
+
+
+def Hex_to_RGB(hex):
+    # 作者的颜色类似于ffffed99 前两位的ff好像没用
+    r = int(hex[2:4], 16) / 255
+    g = int(hex[4:6], 16) / 25
+    b = int(hex[6:8], 16) / 255
+    return (r, g, b)
+
+
+def update_pdf(markup_path, update_log):
+    with open(markup_path, 'r', encoding='UTF-8') as f:
+        markups = json.loads(f.read())
+    pdf_file = markups['title'] + '.pdf'
+    last_update = sorted([x['date'] for x in markups['markups']])[-1]
+
+    if pdf_file not in update_log or last_update > update_log[pdf_file]:
+        doc = fitz.open(
+            os.path.join(os.path.dirname(markup_path), 'resources', pdf_file))
+        for markup in markups['markups']:
+            annot = None
+            page = doc[markup['page']]
+            # 高亮和下划线
+            if markup['type'] == 5:
+                stext = markup['textblocks'][0]['text']
+                rl = page.searchFor(stext, quads=True)
+                while len(rl) == 0 and len(stext) >= 0:
+                    stext = stext[:-1]
+                    rl = page.searchFor(stext, quads=True)
+                # TODO:多行的情况每行新建一条线
+                if len(rl) > 0:
+                    for r in rl:
+                        if 'underline' in markup:
+                            annot = page.addUnderlineAnnot(r)
+                            annot.setColors(
+                                stroke=Hex_to_RGB(markup['linecolor']))
+                            annot.set_border(width=markup['linewidth'])
+                        else:
+                            annot = page.addHighlightAnnot(r)
+                            annot.setColors(
+                                stroke=Hex_to_RGB(markup['fillcolor']))
+                        annot.update()
+            # 打字机和文本框
+            # TODO: 文本框的箭头
+            elif markup['type'] == 9 or markup['type'] == 11:
+                annot = page.addFreetextAnnot(markup['rect'],
+                                              markup['originaltext'],
+                                              text_color=Hex_to_RGB(
+                                                  markup['linecolor']))
+                annot.setColors(stroke=Hex_to_RGB(markup['linecolor']))
+                annot.update()
+            # 方形
+            elif markup['type'] == 2:
+                # TODO: 线宽无法设置
+                annot = page.addRectAnnot(markup['rect'])
+                annot.setColors(stroke=Hex_to_RGB(markup['linecolor']))
+                annot.update()
+            # 圆形
+            elif markup['type'] == 3:
+                # TODO: 线宽无法设置
+                annot = page.addCircleAnnot(markup['rect'])
+                annot.setColors(stroke=Hex_to_RGB(markup['linecolor']))
+                annot.update()
+            # 直线
+            elif markup['type'] == 1:
+                # TODO: 线宽无法设置
+                assert len(markup["rect"]) % 4 == 0
+                for i in range(len(markup["rect"]) // 4):
+                    p1 = (markup["rect"][i * 2 + 0], markup["rect"][i * 2 + 1])
+                    p2 = (markup["rect"][i * 2 + 2], markup["rect"][i * 2 + 3])
+                    annot = page.addLineAnnot(p1, p2)
+                    annot.setColors(stroke=Hex_to_RGB(markup['linecolor']))
+                    annot.set_opacity(0.7)
+                    annot.update()
+            # 其他都没做
+            else:
+                print('Unhandled annotation type:{}'.format(markup['type']))
+            if 'content' in markup and annot is not None:
+                # 下划线、高亮和直线的批注都在最后一个里了
+                info = annot.info
+                info["content"] = markup['content']
+                annot.set_info(info)
+                annot.update()
+        doc.save(os.path.join(dst_pdf_path, pdf_file), deflate=True)
+
+    return pdf_file, last_update
+
+
+if __name__ == "__main__":
+    if not os.path.exists(dst_pdf_path):
+        os.makedirs(dst_pdf_path)
+    # 会在notebooks的目录里生成一个log 仅处理更新过的pdf
+    if os.path.exists(path_join(note_path, 'pdf_log.json')):
+        with open(path_join(note_path, 'pdf_log.json'), 'r') as f:
+            old_update_log = json.load(f)
+    else:
+        old_update_log = {}
+    new_update_log = {}
+
+    notes = glob(path_join(note_path, '*', 'markups.json'))
+    if len(notes) > 0:
+        for idx, markup_path in enumerate(notes):
+            print('{}/{}:\t {}'.format(
+                idx, len(notes),
+                os.path.basename(os.path.dirname(markup_path))))
+            try:
+                pdf_file, last_update = update_pdf(markup_path, old_update_log)
+                new_update_log[pdf_file] = last_update
+            except Exception as e:
+                print(e)
+        # log里只记录本次处理时notepads目录里有的pdf
+        with open(path_join(note_path, 'pdf_log.json'), 'w') as f:
+            json.dump(new_update_log, f)
+    os.system("pause")