@birthday8/doc-mcp 1.0.2 → 1.0.4

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (12) hide show
  1. package/README.md +61 -21
  2. package/index.js +61 -56
  3. package/install.js +45 -35
  4. package/package.json +1 -1
  5. package/python/docx_converter.py +15 -83
  6. package/python/html_rules.py +652 -0
  7. package/python/html_validator.py +59 -274
  8. package/python/html_validator_strict.py +430 -0
  9. package/python/sample/example.html +0 -14
  10. package/python/sample/html_schema.py +352 -0
  11. package/python/server.py +112 -75
  12. package/python/test_strict_validation.py +118 -0
package/README.md CHANGED
@@ -12,7 +12,6 @@ Doc Creator MCP Server - Generate Word documents from HTML with rich formatting
12
12
  - ✅ Tables with styles
13
13
  - ✅ Info/Warning/Success boxes
14
14
  - ✅ Code blocks
15
- - ✅ Blockquotes
16
15
  - ✅ Multi-column layout
17
16
  - ✅ Page breaks
18
17
 
@@ -63,8 +62,20 @@ Generate a Word document from HTML content.
63
62
  }
64
63
  ```
65
64
 
66
- ### 3. get_html_template
67
- Get the HTML template with CSS styles.
65
+ ### 3. get_html_constraints
66
+ Get HTML format constraints example with all supported formats.
67
+
68
+ ### 4. get_html_schema
69
+ Get structured HTML format constraints as JSON.
70
+
71
+ ### 5. validate_html
72
+ Validate HTML content against format constraints.
73
+
74
+ ```json
75
+ {
76
+ "html_content": "<h1>Title</h1><p>Content...</p>"
77
+ }
78
+ ```
68
79
 
69
80
  ## HTML Conventions
70
81
 
@@ -75,12 +86,9 @@ Get the HTML template with CSS styles.
75
86
  <head>
76
87
  <meta charset="UTF-8">
77
88
  <title>Document Title</title>
78
- <style>
79
- /* CSS styles */
80
- </style>
81
89
  </head>
82
90
  <body>
83
- <h1>Title</h1>
91
+ <h1 style="font-family: 黑体; font-size: 18pt; text-align: center;">Title</h1>
84
92
  <p>Content with <strong>bold</strong> and <em>italic</em>.</p>
85
93
  </body>
86
94
  </html>
@@ -92,8 +100,8 @@ Get the HTML template with CSS styles.
92
100
  <em>Italic</em>
93
101
  <u>Underline</u>
94
102
  <s>Strikethrough</s>
95
- <span class="red">Red text</span>
96
- <span class="highlight">Highlighted</span>
103
+ <span style="color: #FF0000;">Red text</span>
104
+ <span style="background-color: #FFFF00;">Highlighted</span>
97
105
  ```
98
106
 
99
107
  ### Paragraph Indentation
@@ -110,17 +118,13 @@ Get the HTML template with CSS styles.
110
118
 
111
119
  ### Tables
112
120
  ```html
113
- <table style="width: 100%; border-collapse: collapse;">
114
- <thead>
115
- <tr>
116
- <th style="border: 1px solid #ddd; padding: 12px;">Header</th>
117
- </tr>
118
- </thead>
119
- <tbody>
120
- <tr>
121
- <td style="border: 1px solid #ddd; padding: 12px;">Cell</td>
122
- </tr>
123
- </tbody>
121
+ <table>
122
+ <tr>
123
+ <th style="background-color: #E3F2FD;">Header</th>
124
+ </tr>
125
+ <tr>
126
+ <td>Cell</td>
127
+ </tr>
124
128
  </table>
125
129
  ```
126
130
 
@@ -130,8 +134,44 @@ Get the HTML template with CSS styles.
130
134
  <p>Column 1 content...</p>
131
135
  <p>Column 2 content...</p>
132
136
  </div>
137
+ <!-- 必须恢复单栏 -->
138
+ <div class="columns" data-cols="1"></div>
139
+ ```
140
+
141
+ ## Important Notes
142
+
143
+ ### Style Usage
144
+ - **Inline styles only**: Use `style="..."` attributes on elements
145
+ - **No `<style>` tags**: CSS in `<style>` tags is not supported
146
+ - **Supported styles**: color, background-color, font-family, font-size, text-align, line-height, margin-top, margin-bottom
147
+ - **Color format**: Must use hex format `#RRGGBB` (e.g., `#FF0000`), NOT `red`, `rgb(255,0,0)`, etc.
148
+ - **Font size format**: Must use `pt` units (e.g., `14pt`), NOT `px`, `em`, `rem`
149
+ - **Margin format**: Must use `pt` units (e.g., `10pt`), NOT `px`, `em`, `rem`
150
+ - **Line height format**: Must be numeric (e.g., `1.5`, `1.8`, `2.0`)
151
+
152
+ ### Class Names
153
+ **Only 7 class names are supported**:
154
+ - `center` - Center aligned paragraph
155
+ - `right` - Right aligned paragraph
156
+ - `left` - Left aligned paragraph
157
+ - `info` - Info message box
158
+ - `warning` - Warning message box
159
+ - `success` - Success message box
160
+ - `columns` - Multi-column layout
161
+
162
+ All other styles must use inline `style` attributes.
163
+
164
+ ### Example with inline styles
165
+ ```html
166
+ <h1 style="font-family: 黑体; font-size: 18pt; text-align: center;">Title</h1>
167
+ <p style="color: #FF0000;">Red text</p>
168
+ <span style="background-color: #FFFF00;">Highlighted</span>
169
+
170
+ <!-- 不支持的类名示例 -->
171
+ <!-- ❌ <span class="abstract-title">摘要</span> -->
172
+ <!-- ✅ <span style="font-family: 黑体; font-size: 12pt; font-weight: bold;">摘要</span> -->
133
173
  ```
134
174
 
135
175
  ## License
136
176
 
137
- MIT
177
+ MIT
package/index.js CHANGED
@@ -5,48 +5,50 @@
5
5
  * 通过npx运行Python MCP Server
6
6
  */
7
7
 
8
- const { spawn } = require('child_process');
9
- const path = require('path');
10
- const fs = require('fs');
8
+ const { spawn } = require("child_process");
9
+ const path = require("path");
10
+ const fs = require("fs");
11
11
 
12
12
  // Python脚本路径
13
- const pythonDir = path.join(__dirname, 'python');
14
- const serverScript = path.join(pythonDir, 'server.py');
15
- const requirementsFile = path.join(pythonDir, 'requirements.txt');
13
+ const pythonDir = path.join(__dirname, "python");
14
+ const serverScript = path.join(pythonDir, "server.py");
15
+ const requirementsFile = path.join(pythonDir, "requirements.txt");
16
16
 
17
17
  // 检查Python是否安装
18
18
  function checkPython() {
19
19
  return new Promise((resolve, reject) => {
20
- const pythonCommands = ['python3', 'python', 'py'];
20
+ const pythonCommands = ["python3", "python", "py"];
21
21
  let found = false;
22
-
22
+
23
23
  function tryNext(index) {
24
24
  if (index >= pythonCommands.length) {
25
- reject(new Error('Python not found. Please install Python 3.8 or higher.'));
25
+ reject(
26
+ new Error("Python not found. Please install Python 3.8 or higher."),
27
+ );
26
28
  return;
27
29
  }
28
-
30
+
29
31
  const cmd = pythonCommands[index];
30
- const check = spawn(cmd, ['--version'], { shell: true });
31
-
32
- check.on('error', () => {
32
+ const check = spawn(cmd, ["--version"], { shell: true });
33
+
34
+ check.on("error", () => {
33
35
  tryNext(index + 1);
34
36
  });
35
-
36
- check.stdout.on('data', (data) => {
37
+
38
+ check.stdout.on("data", (data) => {
37
39
  const version = data.toString().trim();
38
40
  console.error(`[Doc Creator] Found Python: ${version}`);
39
41
  found = true;
40
42
  resolve(cmd);
41
43
  });
42
-
43
- check.on('close', (code) => {
44
+
45
+ check.on("close", (code) => {
44
46
  if (!found) {
45
47
  tryNext(index + 1);
46
48
  }
47
49
  });
48
50
  }
49
-
51
+
50
52
  tryNext(0);
51
53
  });
52
54
  }
@@ -54,16 +56,16 @@ function checkPython() {
54
56
  // 检查依赖是否已安装
55
57
  function checkDependencies(pythonCmd) {
56
58
  return new Promise((resolve, reject) => {
57
- const check = spawn(pythonCmd, ['-c', 'import mcp, docx, bs4'], {
59
+ const check = spawn(pythonCmd, ["-c", "import mcp, docx, bs4"], {
58
60
  cwd: pythonDir,
59
- shell: true
61
+ shell: true,
60
62
  });
61
-
62
- check.on('close', (code) => {
63
+
64
+ check.on("close", (code) => {
63
65
  resolve(code === 0);
64
66
  });
65
-
66
- check.on('error', () => {
67
+
68
+ check.on("error", () => {
67
69
  resolve(false);
68
70
  });
69
71
  });
@@ -72,25 +74,29 @@ function checkDependencies(pythonCmd) {
72
74
  // 安装Python依赖
73
75
  function installDependencies(pythonCmd) {
74
76
  return new Promise((resolve, reject) => {
75
- console.error('[Doc Creator] Installing Python dependencies...');
76
-
77
- const install = spawn(pythonCmd, ['-m', 'pip', 'install', '-r', requirementsFile], {
78
- cwd: pythonDir,
79
- shell: true,
80
- stdio: 'pipe'
81
- });
82
-
83
- install.on('close', (code) => {
77
+ console.error("[Doc Creator] Installing Python dependencies...");
78
+
79
+ const install = spawn(
80
+ pythonCmd,
81
+ ["-m", "pip", "install", "-r", requirementsFile],
82
+ {
83
+ cwd: pythonDir,
84
+ shell: true,
85
+ stdio: "pipe",
86
+ },
87
+ );
88
+
89
+ install.on("close", (code) => {
84
90
  if (code === 0) {
85
- console.error('[Doc Creator] Dependencies installed successfully.');
91
+ console.error("[Doc Creator] Dependencies installed successfully.");
86
92
  resolve();
87
93
  } else {
88
- reject(new Error('Failed to install dependencies'));
94
+ reject(new Error("Failed to install dependencies"));
89
95
  }
90
96
  });
91
-
92
- install.on('error', (err) => {
93
- reject(new Error('Failed to run pip'));
97
+
98
+ install.on("error", (err) => {
99
+ reject(new Error("Failed to run pip"));
94
100
  });
95
101
  });
96
102
  }
@@ -100,46 +106,45 @@ async function startServer() {
100
106
  try {
101
107
  // 检查Python
102
108
  const pythonCmd = await checkPython();
103
-
109
+
104
110
  // 检查依赖
105
111
  const depsInstalled = await checkDependencies(pythonCmd);
106
-
112
+
107
113
  // 如果依赖未安装，尝试安装
108
114
  if (!depsInstalled) {
109
115
  try {
110
116
  await installDependencies(pythonCmd);
111
117
  } catch (err) {
112
118
  console.error(`[Doc Creator] Warning: ${err.message}`);
113
- console.error('[Doc Creator] Trying to continue anyway...');
119
+ console.error("[Doc Creator] Trying to continue anyway...");
114
120
  }
115
121
  }
116
-
122
+
117
123
  // 启动Python MCP Server
118
- console.error('[Doc Creator] Starting MCP Server...');
119
-
124
+ console.error("[Doc Creator] Starting MCP Server...");
125
+
120
126
  const server = spawn(pythonCmd, [serverScript], {
121
127
  cwd: pythonDir,
122
- stdio: ['inherit', 'inherit', 'inherit']
128
+ stdio: ["inherit", "inherit", "inherit"],
123
129
  });
124
-
125
- server.on('error', (err) => {
130
+
131
+ server.on("error", (err) => {
126
132
  console.error(`[Doc Creator] Failed to start server: ${err.message}`);
127
133
  process.exit(1);
128
134
  });
129
-
130
- server.on('close', (code) => {
135
+
136
+ server.on("close", (code) => {
131
137
  process.exit(code);
132
138
  });
133
-
139
+
134
140
  // 处理进程信号
135
- process.on('SIGINT', () => {
136
- server.kill('SIGINT');
141
+ process.on("SIGINT", () => {
142
+ server.kill("SIGINT");
137
143
  });
138
-
139
- process.on('SIGTERM', () => {
140
- server.kill('SIGTERM');
144
+
145
+ process.on("SIGTERM", () => {
146
+ server.kill("SIGTERM");
141
147
  });
142
-
143
148
  } catch (err) {
144
149
  console.error(`[Doc Creator] Error: ${err.message}`);
145
150
  process.exit(1);
package/install.js CHANGED
@@ -5,46 +5,48 @@
5
5
  * 在安装npm包时自动安装Python依赖
6
6
  */
7
7
 
8
- const { spawn } = require('child_process');
9
- const path = require('path');
8
+ const { spawn } = require("child_process");
9
+ const path = require("path");
10
10
 
11
- const pythonDir = path.join(__dirname, 'python');
12
- const requirementsFile = path.join(pythonDir, 'requirements.txt');
11
+ const pythonDir = path.join(__dirname, "python");
12
+ const requirementsFile = path.join(pythonDir, "requirements.txt");
13
13
 
14
14
  // 检查Python是否安装
15
15
  function checkPython() {
16
16
  return new Promise((resolve, reject) => {
17
- const pythonCommands = ['python3', 'python', 'py'];
17
+ const pythonCommands = ["python3", "python", "py"];
18
18
  let found = false;
19
-
19
+
20
20
  function tryNext(index) {
21
21
  if (index >= pythonCommands.length) {
22
- console.log('[Doc Creator] Python not found. Will check again at runtime.');
22
+ console.log(
23
+ "[Doc Creator] Python not found. Will check again at runtime.",
24
+ );
23
25
  resolve(null);
24
26
  return;
25
27
  }
26
-
28
+
27
29
  const cmd = pythonCommands[index];
28
- const check = spawn(cmd, ['--version'], { shell: true });
29
-
30
- check.on('error', () => {
30
+ const check = spawn(cmd, ["--version"], { shell: true });
31
+
32
+ check.on("error", () => {
31
33
  tryNext(index + 1);
32
34
  });
33
-
34
- check.stdout.on('data', () => {
35
+
36
+ check.stdout.on("data", () => {
35
37
  if (!found) {
36
38
  found = true;
37
39
  resolve(cmd);
38
40
  }
39
41
  });
40
-
41
- check.on('close', () => {
42
+
43
+ check.on("close", () => {
42
44
  if (!found) {
43
45
  tryNext(index + 1);
44
46
  }
45
47
  });
46
48
  }
47
-
49
+
48
50
  tryNext(0);
49
51
  });
50
52
  }
@@ -52,26 +54,34 @@ function checkPython() {
52
54
  // 安装Python依赖
53
55
  function installDependencies(pythonCmd) {
54
56
  return new Promise((resolve, reject) => {
55
- console.log('[Doc Creator] Installing Python dependencies...');
56
-
57
- const install = spawn(pythonCmd, ['-m', 'pip', 'install', '-r', requirementsFile], {
58
- cwd: pythonDir,
59
- shell: true,
60
- stdio: 'inherit'
61
- });
62
-
63
- install.on('close', (code) => {
57
+ console.log("[Doc Creator] Installing Python dependencies...");
58
+
59
+ const install = spawn(
60
+ pythonCmd,
61
+ ["-m", "pip", "install", "-r", requirementsFile],
62
+ {
63
+ cwd: pythonDir,
64
+ shell: true,
65
+ stdio: "inherit",
66
+ },
67
+ );
68
+
69
+ install.on("close", (code) => {
64
70
  if (code === 0) {
65
- console.log('[Doc Creator] Python dependencies installed successfully.');
71
+ console.log(
72
+ "[Doc Creator] Python dependencies installed successfully.",
73
+ );
66
74
  resolve();
67
75
  } else {
68
- console.log('[Doc Creator] Failed to install dependencies. Will retry at runtime.');
76
+ console.log(
77
+ "[Doc Creator] Failed to install dependencies. Will retry at runtime.",
78
+ );
69
79
  resolve(); // 不中断安装流程
70
80
  }
71
81
  });
72
-
73
- install.on('error', () => {
74
- console.log('[Doc Creator] Failed to run pip. Will retry at runtime.');
82
+
83
+ install.on("error", () => {
84
+ console.log("[Doc Creator] Failed to run pip. Will retry at runtime.");
75
85
  resolve(); // 不中断安装流程
76
86
  });
77
87
  });
@@ -79,15 +89,15 @@ function installDependencies(pythonCmd) {
79
89
 
80
90
  // 主函数
81
91
  async function main() {
82
- console.log('[Doc Creator] Running post-install script...');
83
-
92
+ console.log("[Doc Creator] Running post-install script...");
93
+
84
94
  const pythonCmd = await checkPython();
85
-
95
+
86
96
  if (pythonCmd) {
87
97
  await installDependencies(pythonCmd);
88
98
  }
89
-
90
- console.log('[Doc Creator] Post-install complete.');
99
+
100
+ console.log("[Doc Creator] Post-install complete.");
91
101
  }
92
102
 
93
103
  main().catch(() => {
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@birthday8/doc-mcp",
3
- "version": "1.0.2",
3
+ "version": "1.0.4",
4
4
  "description": "Doc Creator MCP Server - Generate Word documents from HTML",
5
5
  "main": "index.js",
6
6
  "bin": {
package/python/docx_converter.py CHANGED
@@ -204,11 +204,6 @@ def latex_to_unicode_formula(latex_formula):
204
204
  def add_native_formula(
205
205
  para,
206
206
  latex_formula,
207
- color=None,
208
- font_name=None,
209
- font_size=None,
210
- bold=False,
211
- italic=False,
212
207
  ):
213
208
  """添加 Word 原生公式"""
214
209
  if HAS_MATH2DOCX:
@@ -236,6 +231,7 @@ def parse_color(color_str):
236
231
  return RGBColor(r, g, b)
237
232
  except:
238
233
  return None
234
+ raise
239
235
 
240
236
 
241
237
  def set_font(
@@ -399,11 +395,8 @@ def process_paragraph(
399
395
  para.paragraph_format.line_spacing = para_size * float(line_height_str[:-2])
400
396
  else:
401
397
  # 尝试作为倍数处理
402
- try:
403
- line_spacing_value = float(line_height_str)
404
- para.paragraph_format.line_spacing = line_spacing_value
405
- except:
406
- pass
398
+ line_spacing_value = float(line_height_str)
399
+ para.paragraph_format.line_spacing = line_spacing_value
407
400
 
408
401
  # 解析段后距
409
402
  margin_bottom_match = MARGIN_RE.search(style)
@@ -421,10 +414,7 @@ def process_paragraph(
421
414
  )
422
415
  else:
423
416
  # 尝试作为pt处理
424
- try:
425
- para.paragraph_format.space_after = Pt(float(margin_bottom_str))
426
- except:
427
- pass
417
+ para.paragraph_format.space_after = Pt(float(margin_bottom_str))
428
418
 
429
419
  # 解析段前距
430
420
  margin_top_match = MARGIN_RE.search(style)
@@ -442,10 +432,7 @@ def process_paragraph(
442
432
  )
443
433
  else:
444
434
  # 尝试作为pt处理
445
- try:
446
- para.paragraph_format.space_before = Pt(float(margin_top_str))
447
- except:
448
- pass
435
+ para.paragraph_format.space_before = Pt(float(margin_top_str))
449
436
 
450
437
  # 解析段落的字号
451
438
  para_size = default_size
@@ -464,10 +451,7 @@ def process_paragraph(
464
451
  para_size = float(size_str[:-2]) * 12
465
452
  else:
466
453
  # 尝试直接解析为数字
467
- try:
468
- para_size = float(size_str)
469
- except:
470
- pass
454
+ para_size = float(size_str)
471
455
 
472
456
  # 处理段落内容 - 递归处理所有子元素
473
457
  _process_element_to_runs(paragraph, para, default_font, para_size)
@@ -713,10 +697,7 @@ def _process_element_to_runs(
713
697
  span_size = float(size_str[:-2]) * 12
714
698
  else:
715
699
  # 尝试直接解析为数字
716
- try:
717
- span_size = float(size_str)
718
- except:
719
- pass
700
+ span_size = float(size_str)
720
701
 
721
702
  # 解析class中的颜色
722
703
  class_set = set(classes) # 转换为集合提高查找性能
@@ -1067,7 +1048,7 @@ def _process_heading(element, doc, default_font):
1067
1048
  heading.alignment = WD_PARAGRAPH_ALIGNMENT.CENTER
1068
1049
 
1069
1050
 
1070
- def _process_paragraph_element(element, doc, default_font, default_size):
1051
+ def _process_paragraph_element(element, doc):
1071
1052
  """处理段落元素"""
1072
1053
  classes = element.get("class", [])
1073
1054
  class_set = set(classes)
@@ -1124,17 +1105,12 @@ def _process_image_element(element, doc, html_file):
1124
1105
  width_inch = None
1125
1106
  height_inch = None
1126
1107
  if width:
1127
- try:
1128
- width_px = float(width)
1129
- width_inch = width_px / 96 # 假设96 DPI
1130
- except:
1131
- pass
1108
+ width_px = float(width)
1109
+ width_inch = width_px / 96 # 假设96 DPI
1110
+
1132
1111
  if height:
1133
- try:
1134
- height_px = float(height)
1135
- height_inch = height_px / 96
1136
- except:
1137
- pass
1112
+ height_px = float(height)
1113
+ height_inch = height_px / 96
1138
1114
 
1139
1115
  # 处理相对路径（相对于HTML文件）
1140
1116
  html_dir = os.path.dirname(html_file)
@@ -1251,7 +1227,7 @@ def _process_elements(soup, doc, html_file, default_font, default_size):
1251
1227
  if element.name in ["h1", "h2", "h3", "h4", "h5", "h6"]:
1252
1228
  _process_heading(element, doc, default_font)
1253
1229
  elif element.name == "p":
1254
- _process_paragraph_element(element, doc, default_font, default_size)
1230
+ _process_paragraph_element(element, doc)
1255
1231
  elif element.name == "ul":
1256
1232
  _process_list_element(element, doc, ordered=False)
1257
1233
  elif element.name == "ol":
@@ -1384,51 +1360,6 @@ def convert_html_to_docx(
1384
1360
  elif element.name == "table":
1385
1361
  process_table(element, doc)
1386
1362
 
1387
- elif element.name == "img":
1388
- # 处理图片
1389
- src = element.get("src", "")
1390
- alt = element.get("alt", "")
1391
-
1392
- if src:
1393
- # 解析宽度、高度和对齐方式
1394
- width = element.get("width")
1395
- height = element.get("height")
1396
- style = element.get("style", "")
1397
- align = element.get("align", "center")
1398
-
1399
- # 从 style 中提取对齐方式
1400
- if "text-align: right" in style or "float: right" in style:
1401
- align = "right"
1402
- elif "text-align: left" in style or "float: left" in style:
1403
- align = "left"
1404
- elif "text-align: center" in style:
1405
- align = "center"
1406
-
1407
- # 处理宽度高度（支持像素转英寸）
1408
- width_inch = None
1409
- height_inch = None
1410
- if width:
1411
- try:
1412
- width_px = float(width)
1413
- width_inch = width_px / 96 # 假设96 DPI
1414
- except:
1415
- pass
1416
- if height:
1417
- try:
1418
- height_px = float(height)
1419
- height_inch = height_px / 96
1420
- except:
1421
- pass
1422
-
1423
- # 处理相对路径（相对于HTML文件）
1424
- html_dir = os.path.dirname(html_file)
1425
- image_path = (
1426
- os.path.join(html_dir, src) if not os.path.isabs(src) else src
1427
- )
1428
-
1429
- # 添加图片
1430
- add_image(doc, image_path, width_inch, height_inch, align)
1431
-
1432
1363
  elif element.name == "div":
1433
1364
  # 检查是否是特殊div
1434
1365
  classes = element.get("class", [])
@@ -1578,6 +1509,7 @@ def convert_html_to_docx(
1578
1509
  except:
1579
1510
  para = doc.add_paragraph(f"[图片: {alt}]")
1580
1511
  para.alignment = WD_PARAGRAPH_ALIGNMENT.CENTER
1512
+ raise
1581
1513
  else:
1582
1514
  para = doc.add_paragraph(f"[图片: {alt} - 路径: {src}]")
1583
1515
  para.alignment = WD_PARAGRAPH_ALIGNMENT.CENTER