webMagic-Mobai
/
magic-Dependency
mirror of https://gitea.qinmian.online/Mobai/magic-Dependency.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #556 from zhuyuesut/master 

增加对零宽断言的支持
master
Yihua Huang 2017-06-03 11:10:56 +08:00 committed by GitHub
parent f02f469c69 9e1b7ed3f7
commit b6b991a09b
2 changed files with 10 additions and 32 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

38
webmagic-core/src/main/java/us/codecraft/webmagic/selector/RegexSelector.java
View File

@ -26,14 +26,16 @@ public class RegexSelector implements Selector {
if (StringUtils.isBlank(regexStr)) {
throw new IllegalArgumentException("regex must not be empty");
}
// Check bracket for regex group. Add default group 1 if there is no group.
// Only check if there exists the valid left parenthesis, leave regexp validation for Pattern.
if ( ! hasGroup(regexStr) ){
regexStr = "(" + regexStr + ")";
}
this.regexStr = regexStr;
try {
regex = Pattern.compile(regexStr, Pattern.DOTALL | Pattern.CASE_INSENSITIVE);
// Check bracket for regex group. Add default group 1 if there is no group.
// Only check if there exists the valid left parenthesis, leave regexp validation for Pattern.
if ( regex.matcher("").groupCount() == 0 ){
regexStr = "(" + regexStr + ")";
regex = Pattern.compile(regexStr, Pattern.DOTALL | Pattern.CASE_INSENSITIVE);
}
this.regexStr = regexStr;
} catch (PatternSyntaxException e) {
throw new IllegalArgumentException("invalid regex", e);
}
@ -44,30 +46,6 @@ public class RegexSelector implements Selector {
this(regexStr, 1);
}
private boolean hasGroup(String regexStr) {
if (StringUtils.countMatches(regexStr, "(") - StringUtils.countMatches(regexStr, "\\(") ==
StringUtils.countMatches(regexStr, "(?:") - StringUtils.countMatches(regexStr, "\\(?:")){
return false;
}
if (StringUtils.countMatches(regexStr, "(") - StringUtils.countMatches(regexStr, "\\(") ==
StringUtils.countMatches(regexStr, "(?=") - StringUtils.countMatches(regexStr, "\\(?=") ) {
return false;
}
if (StringUtils.countMatches(regexStr, "(") - StringUtils.countMatches(regexStr, "\\(") ==
StringUtils.countMatches(regexStr, "(?<") - StringUtils.countMatches(regexStr, "\\(?<") ) {
return false;
}
if (StringUtils.countMatches(regexStr, "(") - StringUtils.countMatches(regexStr, "\\(") ==
StringUtils.countMatches(regexStr, "(?!") - StringUtils.countMatches(regexStr, "\\(?!") ) {
return false;
}
if (StringUtils.countMatches(regexStr, "(") - StringUtils.countMatches(regexStr, "\\(") ==
StringUtils.countMatches(regexStr, "(?#") - StringUtils.countMatches(regexStr, "\\(?#") ) {
return false;
}
return true;
}
@Override
public String select(String text) {
return selectGroup(text).get(group);

4
webmagic-core/src/test/java/us/codecraft/webmagic/selector/RegexSelectorTest.java
View File

@ -25,8 +25,8 @@ public class RegexSelectorTest {
@Test
public void testRegexWithZeroWidthAssertions() {
String regex = "^.*(?=\\?)";
String source = "hello world?xxxx";
String regex = "^.*(?=\\?)(?!\\?yy)";
String source = "hello world?xx?yy";
RegexSelector regexSelector = new RegexSelector(regex);
String select = regexSelector.select(source);
Assertions.assertThat(select).isEqualTo("hello world");