Skip to content
项目
群组
代码片段
帮助
当前项目
正在载入...
登录 / 注册
切换导航面板
F
fuse
Project
Project
详情
活动
周期分析
仓库
仓库
文件
提交
分支
标签
贡献者
Graph
比较
统计图
议题
0
议题
0
列表
看板
标记
Milestones
合并请求
0
合并请求
0
CI / CD
CI / CD
流水线
作业
日程
统计图
注册表
注册表
Wiki
Wiki
代码片段
代码片段
成员
成员
折叠边栏
关闭边栏
活动
图像
聊天
创建新问题
作业
提交
问题看板
Open sidebar
任建彩
fuse
Commits
4bd7b536
提交
4bd7b536
authored
3月 21, 2023
作者:
任建彩
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
配置
上级
fc80d1ca
隐藏空白字符变更
内嵌
并排
正在显示
1 个修改的文件
包含
83 行增加
和
31 行删除
+83
-31
WordController.java
...java/com/zrqx/file/controller/backend/WordController.java
+83
-31
没有找到文件。
szzy-provider/szzy-provider-file/src/main/java/com/zrqx/file/controller/backend/WordController.java
浏览文件 @
4bd7b536
...
...
@@ -21,6 +21,7 @@ import java.io.File;
import
java.io.IOException
;
import
java.util.ArrayList
;
import
java.util.List
;
import
java.util.regex.Matcher
;
import
java.util.regex.Pattern
;
/**
...
...
@@ -41,8 +42,8 @@ public class WordController {
* @return
*/
@ApiOperation
(
value
=
"解析文件目录"
)
@RequestMapping
(
value
=
"/upload"
,
method
=
RequestMethod
.
POST
)
@ApiOperation
(
value
=
"解析文件目录
1
"
)
@RequestMapping
(
value
=
"/upload
1
"
,
method
=
RequestMethod
.
POST
)
public
CallBack
<
List
<
String
>>
searchWordDocX
(
@RequestParam
(
"fileFolder"
)
MultipartFile
file
)
throws
IOException
{
FileInfo
entity
=
service
.
uploadFile
(
file
);
String
fileName
=
entity
.
getOriginalFileName
();
...
...
@@ -56,40 +57,57 @@ public class WordController {
}
return
CallBack
.
success
(
list
);
}
/*public List<String> fileWord(String filePath){
@ApiOperation
(
value
=
"解析文件目录2"
)
@RequestMapping
(
value
=
"/upload2"
,
method
=
RequestMethod
.
POST
)
public
CallBack
<
List
<
String
>>
searchWordDoc2
(
@RequestParam
(
"fileFolder"
)
MultipartFile
file
)
throws
IOException
{
FileInfo
entity
=
service
.
uploadFile
(
file
);
String
fileName
=
entity
.
getOriginalFileName
();
String
suff
=
fileName
.
substring
(
fileName
.
lastIndexOf
(
"."
)
+
1
);
String
content
=
""
;
String
filePath
=
null
;
filePath
=
rootPath
+
entity
.
getPath
()+
"/"
+
entity
.
getFileName
()+
entity
.
getSuffixName
();
List
<
String
>
list
=
new
ArrayList
<>();
Document doc = new Document();
doc.loadFromFile(filePath);
//获取段落数量
int count = doc.getSections().get(0).getParagraphs().getCount();
System.out.println("总共含有段落数:" + count);
//获取段落
int i=0;
int j=0;
for (int z = 0; z < count; z++) {
Paragraph paragraph = doc.getSections().get(0).getParagraphs().get(z);
if(StringUtils.isNotBlank(paragraph.getText())){
//获取子段落
String text = paragraph.getText();
String substring = text.substring(0, 1);
if(substring.contains("(")){
int mid = z - 1;
System.out.println("获取二级段落:<p id='p"+j+"' pid='"+i+"'>"+paragraph.getText()+"</p>");
content="获取二级段落:<p id='p"+j+"' pid='"+i+"'>"+paragraph.getText()+"</p>";
list.add(content);
}else{
i++;
System.out.println("获取一级段落:<p id='"+i+"'>"+paragraph.getText()+"</p>");
content="获取一级段落:<p id='"+i+"'>"+paragraph.getText()+"</p>";
list.add(content);
if
(
suff
.
equals
(
"html"
))
{
File
file1
=
new
File
(
filePath
);
org
.
jsoup
.
nodes
.
Document
doc
=
Jsoup
.
parse
(
file1
,
"UTF-8"
);
//读取.html文件为字符串
Elements
es12
=
doc
.
getElementsByClass
(
"msgdet_left_con"
);
for
(
Element
d:
es12
)
{
Elements
p1
=
d
.
getElementsByTag
(
"p"
);
if
(
p1
.
size
()>
0
){
for
(
Element
el1:
p1
)
{
//获取a标签下的文本内容
String
text
=
el1
.
text
();
if
(
StringUtils
.
isNotBlank
(
text
))
{
String
regEx
=
"[^0-9]"
;
Pattern
p
=
Pattern
.
compile
(
regEx
);
Matcher
m
=
p
.
matcher
(
text
);
String
result
=
m
.
replaceAll
(
""
).
trim
();
int
length
=
result
.
length
();
if
(
length
==
3
){
System
.
out
.
println
(
"获取三级段落<p class='c03' >"
+
text
+
"</p>"
);
content
=
"获取二级段落<p class='c03'>"
+
text
+
"</p>"
;
list
.
add
(
content
);
}
if
(
length
==
2
){
System
.
out
.
println
(
"获取二级段落<p class='c02' >"
+
text
+
"</p>"
);
content
=
"获取二级段落<p class='c02'>"
+
text
+
"</p>"
;
list
.
add
(
content
);
}
if
(
length
==
1
){
System
.
out
.
println
(
"获取一级段落<p class='c01' >"
+
text
+
"</p>"
);
content
=
"获取一级段落<p class='c01'>"
+
text
+
"</p>"
;
list
.
add
(
content
);
}
}
}
}
}
}
return
list
;
}
*/
return
CallBack
.
success
(
list
)
;
}
public
List
<
String
>
fileHtml
(
String
filePath
)
throws
IOException
{
// filePath = "D:/opt/upload/rsgw/2023/03/21/28622e72956e49a4ae2b6b79ea871c6a.html";
// filePath = "D:/opt/upload/rsgw/2023/03/21/28622e72956e49a4ae2b6b79ea871c6a.html";
String
content
=
""
;
List
<
String
>
list
=
new
ArrayList
<>();
File
file
=
new
File
(
filePath
);
...
...
@@ -127,7 +145,7 @@ public class WordController {
boolean
hanzi
=
Pattern
.
matches
(
ze
,
substring
);
if
(
substring
.
contains
(
"("
)){
j
++;
if
(
text
.
contains
(
"。"
)){
//是否包含 ?
if
(
text
.
contains
(
"。"
)){
text
=
text
.
substring
(
0
,
text
.
indexOf
(
"。"
));
//拿取字符串 从第一位开始到问号前结束
}
System
.
out
.
println
(
"获取二级段落:<p class='c02'>"
+
text
+
"</p>"
);
...
...
@@ -142,4 +160,38 @@ public class WordController {
return
list
;
}
/*public List<String> fileWord(String filePath){
String content = "";
List<String> list = new ArrayList<>();
Document doc = new Document();
doc.loadFromFile(filePath);
//获取段落数量
int count = doc.getSections().get(0).getParagraphs().getCount();
System.out.println("总共含有段落数:" + count);
//获取段落
int i=0;
int j=0;
for (int z = 0; z < count; z++) {
Paragraph paragraph = doc.getSections().get(0).getParagraphs().get(z);
if(StringUtils.isNotBlank(paragraph.getText())){
//获取子段落
String text = paragraph.getText();
String substring = text.substring(0, 1);
if(substring.contains("(")){
int mid = z - 1;
System.out.println("获取二级段落:<p id='p"+j+"' pid='"+i+"'>"+paragraph.getText()+"</p>");
content="获取二级段落:<p id='p"+j+"' pid='"+i+"'>"+paragraph.getText()+"</p>";
list.add(content);
}else{
i++;
System.out.println("获取一级段落:<p id='"+i+"'>"+paragraph.getText()+"</p>");
content="获取一级段落:<p id='"+i+"'>"+paragraph.getText()+"</p>";
list.add(content);
}
}
}
return list;
}*/
}
编写
预览
Markdown
格式
0%
重试
或
添加新文件
添加附件
取消
您添加了
0
人
到此讨论。请谨慎行事。
请先完成此评论的编辑!
取消
请
注册
或者
登录
后发表评论