后台处理文档

AIProofread/Bridge.cs

@@ -1,4 +1,5 @@
 using AIProofread.Controls;
+using AIProofread.Util;
 using Microsoft.Office.Interop.Word;
 using Microsoft.Office.Tools.Word;
 using Microsoft.Web.WebView2.Core;
@@ -11,8 +12,10 @@ using System.Drawing;
 using System.IO;
 using System.Runtime.InteropServices;
 using System.Text.RegularExpressions;
+using System.Threading.Tasks;
 using UtilLib;
 using Document = Microsoft.Office.Interop.Word.Document;
+using Task = System.Threading.Tasks.Task;
 
 namespace AIProofread
 {
@@ -149,21 +152,114 @@ namespace AIProofread
         // 获取文档所有文本数据
         public Dictionary<string, object> getAllText()
         {
-            return Tools.GetAllText();
+            return Tools.GetAllText(Globals.ThisAddIn.Application.ActiveDocument);
         }
         public string getDocumentData()
         {
             Dictionary<string, object> data = new Dictionary<string, object>();
-            var name = Globals.ThisAddIn.Application.ActiveDocument.Name;
             var doc = Globals.ThisAddIn.Application.ActiveDocument;
-            data.Add("name", name);
+            data.Add("name", doc.Name);
             data.Add("fullName", doc.FullName);
             data.Add("wordsCount", doc.Words.Count);
             data.Add("charactersCount", doc.Characters.Count);
-            data.Add("content", Tools.GetAllText());
+
+            Application app = new Application();
+
+            object missing = System.Reflection.Missing.Value;
+            object saveOption = WdSaveOptions.wdDoNotSaveChanges;
+            // 创建临时文件
+            string path = CreateTempDocumentFile(doc);
+            Document tmpDoc = app.Documents.Open(path, false, true, false);
+            data.Add("content", Tools.GetAllText(tmpDoc));
+            // 关闭并释放对象
+            tmpDoc.Close(ref saveOption, ref missing, ref missing);
+            Marshal.ReleaseComObject(tmpDoc);
+            app.Quit(ref saveOption,ref missing,ref missing);
+            Marshal.ReleaseComObject(app);
+            // 显式回收一下吧
+            GC.Collect();
+            File.Delete(path);
             return Tools.GetJSONString(data);
         }
 
+        private string CreateTempDocumentFile(Document doc)
+        {
+            // 先转成临时文件
+            FileStream fs = new FileStream(doc.FullName, FileMode.Open, FileAccess.Read, FileShare.ReadWrite);
+            // 
+            var tmpFile = Path.GetTempFileName();
+            byte[] bytes = new byte[fs.Length];
+            fs.Read(bytes, 0, bytes.Length);
+            FileStream tmp = new FileStream(tmpFile, FileMode.OpenOrCreate,FileAccess.ReadWrite);
+            tmp.Write(bytes, 0, bytes.Length);
+            // 关闭吧
+            tmp.Close();
+            fs.Close();
+
+            return tmpFile;
+        }
+
+
+        public string getParagraphTextByRange(int start, int end)
+        {
+            var list = Tools.GetTextListByParagraphRange(start, end);
+            return Tools.GetJSONString(list);
+        }
+
+        public string getAllParagraphs()
+        {
+            var doc = Globals.ThisAddIn.Application.ActiveDocument;
+
+            var rangeText = doc.Content.Text;
+            var trimText = HostHelper.ReplaceSpecialChars(rangeText, isReplaceMultSpaceLine: true);
+
+            string[] separator = new string[5] { "\r\a", "\a", "\r", "\v", "\f" };
+            string[] array4 = rangeText.Split(separator, StringSplitOptions.None);
+            string[] array5 = trimText.Split('\n');
+            List<string> list = new List<string>();
+            var paragraphs = doc.Paragraphs;
+            int total = paragraphs.Count;
+            for (int i = 1; i <= total; i++)
+            {
+                list.Add(GetParagraphText(paragraphs[i]));
+            }
+            var array6 = list.ToArray();
+            Dictionary<string, object> data = new Dictionary<string, object>
+            {
+                {"origin_cut",array4 },
+                {"trim_cut",array5 },
+                {"paragraph_cut",array6 },
+            };
+            return Tools.GetJSONString(data);
+        }
+        private string GetParagraphText(Paragraph paragraph)
+        {
+            // 需要
+            return GetRangeText(paragraph.Range);
+        }
+        private string GetRangeText(Range range)
+        {
+            // 需要
+            return range.Text;
+        }
+
+
+        public void getParagraphTextByRangeSync(int start,int end)
+        {
+            //var list = Tools.GetTextListByParagraphRange(start, end);
+            //return Tools.GetJSONString(list);
+            Task.Run(() =>
+            {
+                var list = Tools.GetTextListByParagraphRange(start, end);
+                Globals.ThisAddIn.SendMessageToWeb("getParagraphTextByRange", Tools.GetJSONString(list));
+            });
+        }
+
+        public int getTotalParagraphNumber()
+        {
+            return Globals.ThisAddIn.Application.ActiveDocument.Paragraphs.Count;
+        }
+
         /// <summary>
         /// 读取文档原始文件并转换成base64
         /// </summary>
@@ -482,18 +578,28 @@ namespace AIProofread
                         int index = 0;
                         foreach (var item in correct.Diffs)
                         {
+                            if (item.idx == 3330)
+                            {
+                                Console.WriteLine("xx");
+                            }
                             var mark = AddBookmark(item, index, correct.Sentence_offset, correct.Insert_len, correct.Paragraph_num);
                             if (item.tag != "i") index++;
-                            var msg = new Dictionary<string, object>{
-                                {"message",mark == null ? "没有找到标记对象":"标记对象" + mark.Name },
-                                { "origin",item }
-                            };
                             if (mark != null)
                             {
                                 marks.Add(item.id, new ProofreadItem(item, mark));
                             }
+                            else
+                            {
+                                var msg = new Dictionary<string, object>{
+                                    {"message","没有找到标记对象" },
+                                    { "origin",item },
+                                    { "origin_correct",correct },
+                                    { "new_text",correct.New_text },
+                                    { "paragraph_num",correct.Paragraph_num },
+                                };
+                                Logger.Log(JsonConvert.SerializeObject(msg));
+                            }
 
-                            Logger.LogToWeb(JsonConvert.SerializeObject(msg));
                         }
                     }
                 }

AIProofread/Util/HostHelper.cs

@@ -0,0 +1,52 @@
+using System;
+using System.Collections.Generic;
+using System.Linq;
+using System.Text.RegularExpressions;
+
+namespace AIProofread.Util
+{
+    public class HostHelper
+    {
+        public static string ReplaceSpecialChars(string text, bool isReplaceMultSpaceLine = false)
+        {
+            Regex regex = new Regex("[\r\f\a\v]{1}");
+            if (!string.IsNullOrEmpty(text))
+            {
+                text = regex.Replace(text, "\n").Replace("\u001e", "-");
+                if (isReplaceMultSpaceLine)
+                {
+                    bool flag = text.Last() == '\n';
+                    text = string.Join("\n", text.Split(new char[1] { '\n' }, StringSplitOptions.RemoveEmptyEntries));
+                    if (flag)
+                    {
+                        text += "\n";
+                    }
+                }
+            }
+
+            return text;
+        }
+
+        public static string RemoveInvisibleCharas(string text, bool containTable = false)
+        {
+            List<char> list = new List<char> { '\r', '\a', '\n', '\f', '\v' };
+            if (containTable)
+            {
+                list.Add('\t');
+            }
+
+            return string.IsNullOrEmpty(text) ? string.Empty : text.Trim(list.ToArray());
+        }
+
+        public static bool HasValidCharacters(string text)
+        {
+            List<char> list = new List<char>{
+                '\0', '\u0001', '\u0002', '\u0003', '\u0004', '\u0005', '\u0006', '\a', '\b', '\t',
+                '\n', '\v', '\f', '\r', '\u000e', '\u000f', '\u0010', '\u0011', '\u0012', '\u0013',
+                '\u0014', '\u0015', '\u0016', '\u0017', '\u0018', '\u0019', '\u001a', '\u001b', '\u001c', '\u001d',
+                '\u001e', '\u001f', ' ', '/'
+            };
+            return text.Trim(list.ToArray()).Any();
+        }
+    }
+}

AIProofread/core/Tools.cs

@@ -1,72 +1,56 @@
-using AIProofread.Controls;
+using Microsoft.Office.Interop.Word;
-using Microsoft.Office.Interop.Word;
 using Newtonsoft.Json;
 using System;
 using System.Collections.Generic;
-using System.Linq;
+using System.Diagnostics;
 using System.Security.Cryptography;
-using System.Security.Policy;
-using System.Text;
-using System.Windows.Forms;
-using static System.Net.Mime.MediaTypeNames;
 
 namespace AIProofread
 {
     public class Tools
     {
         private static readonly string[] paragSplitor = new string[] { "\r", "\n", "\r\n" };
-        public static Dictionary<string, object> GetAllText()
+        public static Dictionary<string, object> GetAllText(Document doc)
         {
-            var doc = Globals.ThisAddIn.Application.ActiveDocument;
             // 获取当前文档所有文本
             string allText = doc.Range().Text;
             List<DocumentText> list = new List<DocumentText>();
             //    // 开始分割
             MD5 md5 = new MD5CryptoServiceProvider();
 
-            //if (allText != null && allText.Trim().Length > 0)
-            //{
-
-            //    List<string> lines = allText.Split(paragSplitor, StringSplitOptions.None).ToList();//StringUtil.CutTextToSentences(allText);
-            //    foreach (string text in lines)
-            //    {
-            //        if(text.Trim().Length > 0)
-            //        {
-            //            byte[] hash = md5.ComputeHash(Encoding.Default.GetBytes(text));
-            //            list.Add(new DocumentText(hash, text + "\n"));
-            //        }
-            //        else
-            //        {
-
-            //            list.Add(new DocumentText(text + "\n"));
-            //        }
-            //    }
-            //}
-
             Paragraphs paragraphs = doc.Paragraphs;
-            int total = paragraphs.Count;
+
             //FormLoading frm = new FormLoading();
             //frm.Show();
-            for (int paragraphNumber = 1; paragraphNumber <= total; paragraphNumber++)
+            int paragraphNumber = 0;
+            foreach (Paragraph p in paragraphs)
             {
-                Paragraph p = paragraphs[paragraphNumber];
+                paragraphNumber++;
+                if (paragraphNumber % 20 == 0)
+                {
+                    Debug.WriteLine("process paragraphNumber{0}", paragraphNumber);
+                }
                 Range r = p.Range;
-                if (r.ListFormat.ListType == WdListType.wdListPictureBullet
+                string text = p.Range.Text;
-                    || r.Tables.Count > 0 
+                if (text.Trim().Length == 0 || text.EndsWith("\r\a") || r.Tables.Count > 0)
-                    || p.Range.Text.Trim().Length == 0)
                 {
                     continue;
                 }
-                string text = p.Range.Text;
+                list.Add(new DocumentText(text.Replace("\u0002", ""), paragraphNumber));
-                //Logger.LogToWeb(string.Format("get paragraph {0}", paragraphNumber));
-                //frm.SetLoadingText(text);
-                if (text.Trim().Length > 0)
-                {
-                    //byte[] hash = md5.ComputeHash(Encoding.Default.GetBytes(text));
-                    list.Add(new DocumentText(text,paragraphNumber));
             }
+            //for (; paragraphNumber <= total; paragraphNumber++)
+            //{
 
-            }
+            //    Paragraph p = paragraphs[paragraphNumber];
+
+            //    //Logger.LogToWeb(string.Format("get paragraph {0}", paragraphNumber));
+            //    //frm.SetLoadingText(text);
+            //    //if (text.Trim().Length > 0)
+            //    //{
+            //    //    //byte[] hash = md5.ComputeHash(Encoding.Default.GetBytes(text));
+
+            //    //}
+            //}
             //frm.Close();
             var map = new Dictionary<string, object>
             {
@@ -76,6 +60,36 @@ namespace AIProofread
             return map;
         }
 
+        public static List<DocumentText> GetTextListByParagraphRange(int start, int end)
+        {
+            List<DocumentText> list = new List<DocumentText>();
+            var doc = Globals.ThisAddIn.Application.ActiveDocument;
+            Paragraphs paragraphs = doc.Paragraphs;
+            int total = paragraphs.Count;
+            if (start > total) return list;
+            start = Math.Max(start, 1);
+            end = Math.Min(end, total);
+
+            for (int paragraphNumber = start; paragraphNumber <= end; paragraphNumber++)
+            {
+                Paragraph p = paragraphs[paragraphNumber];
+                Range r = p.Range;
+                if (r.Tables.Count > 0
+                    || r.Endnotes.Count > 0
+                    || r.Footnotes.Count > 0
+                    || r.ListFormat.ListType != WdListType.wdListNoNumbering)
+                {
+                    continue;
+                }
+                string text = p.Range.Text;
+                if (text.Trim().Length > 0)
+                {
+                    list.Add(new DocumentText(text, paragraphNumber));
+                }
+            }
+            return list;
+        }
+
         public static string GetJSONString(object data)
         {
             return JsonConvert.SerializeObject(data, Formatting.Indented);