parent
29e482943b
commit
0822796201
GPG Key ID:
947ADABD8533C476
@ -0,0 +1,146 @@ |
||||
package cn.linghang.mywust.core.parser.undergraduate.global; |
||||
|
||||
import cn.linghang.mywust.core.exception.ParseException; |
||||
import cn.linghang.mywust.model.global.Classroom; |
||||
import cn.linghang.mywust.model.global.Course; |
||||
import cn.linghang.mywust.util.StringUtil; |
||||
import org.jsoup.Jsoup; |
||||
import org.jsoup.nodes.Element; |
||||
import org.jsoup.select.Elements; |
||||
|
||||
import java.util.ArrayList; |
||||
import java.util.List; |
||||
import java.util.regex.Matcher; |
||||
import java.util.regex.Pattern; |
||||
|
||||
abstract class GlobalCourseTableParser { |
||||
|
||||
private static final Pattern weekPattern = Pattern.compile("\\d+"); |
||||
|
||||
protected List<Course> parse(String html, String xpath, Pattern infoPattern) throws ParseException { |
||||
// 直接一步到位，拿到所有的大格子
|
||||
Elements bigGirds = Jsoup.parse(html).selectXpath(xpath); |
||||
System.out.println("Girds: " + bigGirds.size()); |
||||
if (bigGirds.isEmpty()) { |
||||
return new ArrayList<>(); |
||||
} |
||||
|
||||
List<Course> courses = new ArrayList<>(bigGirds.size()); |
||||
|
||||
int girdCount = 0; |
||||
for (Element bigGird : bigGirds) { |
||||
// 处理完一行，计数器置零，方便后续的计算
|
||||
if (girdCount == 42) { |
||||
girdCount = 0; |
||||
} |
||||
|
||||
// 大格子是空的，没课，跳过就好了
|
||||
if (bigGird.text().length() == 0) { |
||||
girdCount++; |
||||
continue; |
||||
} |
||||
|
||||
// 拿到格子里面所有的课
|
||||
Elements courseGirds = bigGird.getElementsByTag("div"); |
||||
|
||||
// 最左边的格子是表的首列，不是课程，拿到的集合是空的，直接跳过，不要计数，以免影响计算节次和星期信息
|
||||
if (courseGirds.isEmpty()) { |
||||
continue; |
||||
} |
||||
|
||||
// 通过计数器计算上课星期和节次信息
|
||||
int weekDay = girdCount / 6 + 1; |
||||
int startSection = (girdCount % 6) * 2 + 1; |
||||
|
||||
Course.CourseBuilder courseBuilder = Course.builder(); |
||||
courseBuilder.weekDay(weekDay); |
||||
courseBuilder.startSection(startSection); |
||||
courseBuilder.endSection(startSection + 1); |
||||
|
||||
// 解析格子里的课
|
||||
for (Element courseGird : courseGirds) { |
||||
Matcher courseInfoMatcher = infoPattern.matcher(courseGird.ownText()); |
||||
if (!courseInfoMatcher.find()) { |
||||
continue; |
||||
} |
||||
|
||||
courseBuilder.name(courseInfoMatcher.group("name")); |
||||
courseBuilder.teacher(courseInfoMatcher.group("teacher")); |
||||
courseBuilder.teachClass(courseInfoMatcher.group("teachClass")); |
||||
courseBuilder.classroom(this.parseClassroom(courseInfoMatcher.group("building"))); |
||||
|
||||
// 解析周次，不使用String.split而是手动分割，是因为系统自带split方法每次调用都需要编译一次切割正则，这里需要执行次数较多，效率不太行
|
||||
List<String> weeks = StringUtil.split(courseInfoMatcher.group("weekString"), ','); |
||||
for (String week : weeks) { |
||||
Matcher weekMatcher = weekPattern.matcher(week); |
||||
// 周次信息不是数字，这种情况尚未出现过，这里的if判断只是用于消除warming
|
||||
if (!weekMatcher.find()) { |
||||
continue; |
||||
} |
||||
|
||||
int startWeek = Integer.parseInt(weekMatcher.group()); |
||||
// 再执行一次matcher.find()，如果没有数字匹配说明是单周课程
|
||||
int endWeek = weekMatcher.find() ? Integer.parseInt(weekMatcher.group()) : startWeek; |
||||
|
||||
courseBuilder.startWeek(startWeek); |
||||
courseBuilder.endWeek(endWeek); |
||||
|
||||
courses.add(courseBuilder.build()); |
||||
} |
||||
} |
||||
|
||||
// 别忘了这里还有个计数器
|
||||
girdCount++; |
||||
} |
||||
|
||||
return courses; |
||||
} |
||||
|
||||
// 搬运修改自老项目
|
||||
private static final Pattern BUILDING_ROOM_INTERNATION_PATTERN = Pattern.compile("(?<building>.*楼)(?<area>.*区)(?<room>\\d+)\\((国际专用)\\)"); // xx楼x区xxx(国际专用)
|
||||
|
||||
private static final Pattern BUILDING_ROOM_CAMPUS_PATTERN = Pattern.compile("(?<building>.*楼)(?<room>\\d+)\\((?<campus>.*)\\)"); // xx楼xxx(黄家湖)
|
||||
|
||||
private static final Pattern BUILDING_AREA_ROOM_PATTERN = Pattern.compile("(?<building>.*楼)(?<area>.*区)(?<room>\\d+)"); // xx楼x区xxx
|
||||
|
||||
private static final Pattern BUILDING_COLLEGE_ROOM_PATTERN = Pattern.compile("(?<building>.*楼)\\((.*)\\)(?<room>\\d+)"); //xx楼(xxx)xxx
|
||||
|
||||
private static final Pattern BUILDING_ROOM = Pattern.compile("(?<building>.*楼)(?<room>\\d+)"); // xx楼xxx
|
||||
|
||||
/** |
||||
* 解析上课地点 |
||||
* |
||||
* @param placeName 上课地点名称 |
||||
* @return 解析后的上课地点 |
||||
*/ |
||||
protected Classroom parseClassroom(String placeName) { |
||||
// 用正则一个一个格式匹配，看起来有点不太聪明的样子，但是目前并未想到更好的办法...
|
||||
// 下面的顺序是按照频率排好的，概率越高，就放在更前面，减少正则匹配的次数
|
||||
Matcher matcher = BUILDING_AREA_ROOM_PATTERN.matcher(placeName); |
||||
if (matcher.find()) { |
||||
return new Classroom("", matcher.group("building"), matcher.group("area"), matcher.group("room")); |
||||
} |
||||
|
||||
matcher = BUILDING_ROOM.matcher(placeName); |
||||
if (matcher.find()) { |
||||
return new Classroom("", matcher.group("building"), "", matcher.group("room")); |
||||
} |
||||
|
||||
matcher = BUILDING_ROOM_CAMPUS_PATTERN.matcher(placeName); |
||||
if (matcher.find()) { |
||||
return new Classroom(matcher.group("campus"), matcher.group("building"), "", matcher.group("room")); |
||||
} |
||||
|
||||
matcher = BUILDING_COLLEGE_ROOM_PATTERN.matcher(placeName); |
||||
if (matcher.find()) { |
||||
return new Classroom("", matcher.group("building"), "", matcher.group("room")); |
||||
} |
||||
|
||||
matcher = BUILDING_ROOM_INTERNATION_PATTERN.matcher(placeName); |
||||
if (matcher.find()) { |
||||
return new Classroom("", matcher.group("building"), matcher.group("area"), matcher.group("room")); |
||||
} |
||||
|
||||
return new Classroom("", "", "", placeName); |
||||
} |
||||
} |
||||
@ -0,0 +1,22 @@ |
||||
package cn.linghang.mywust.core.parser.undergraduate.global; |
||||
|
||||
import cn.linghang.mywust.core.exception.ParseException; |
||||
import cn.linghang.mywust.core.parser.Parser; |
||||
import cn.linghang.mywust.model.global.Course; |
||||
|
||||
import java.util.List; |
||||
import java.util.regex.Pattern; |
||||
|
||||
|
||||
public class UndergradAllCourseScheduleParser extends GlobalCourseTableParser implements Parser<List<Course>> { |
||||
private static final String courseGirdsXPath = "//*[@id=\"kbtable\"]/tbody/tr/td"; |
||||
|
||||
// name组的如果使用.*?匹配而不是.*贪婪匹配的话虽然可以大大减少匹配次数，但是对于课程名中有空格的课程，可能会导致解析错误，
|
||||
// 不过对于没有老师且课程名中含有空格的课程（都是一些在线课程），虽然能匹配，但是字段会有偏差错误，不过影响不大就是了，在线课程不用太过于纠结
|
||||
private static final Pattern pattern = Pattern.compile("(?<name>.*) (?<teachClass>.*?) (?<teacher>.*?) \\((?<weekString>.*?)周\\) ?(?<building>.*)"); |
||||
|
||||
@Override |
||||
public List<Course> parse(String html) throws ParseException { |
||||
return super.parse(html, courseGirdsXPath, pattern); |
||||
} |
||||
} |
||||
@ -0,0 +1,22 @@ |
||||
package cn.linghang.mywust.core.parser.undergraduate.global; |
||||
|
||||
import cn.linghang.mywust.core.exception.ParseException; |
||||
import cn.linghang.mywust.core.parser.Parser; |
||||
import cn.linghang.mywust.model.global.Course; |
||||
|
||||
import java.util.List; |
||||
import java.util.regex.Pattern; |
||||
|
||||
|
||||
public class UndergradTeacherCourseParser extends GlobalCourseTableParser implements Parser<List<Course>> { |
||||
private static final String courseGirdsXPath = "//*[@id=\"kbtable\"]/tbody/tr/td"; |
||||
|
||||
// 还有教学班名称格式为20xx寒假课堂xx的线上课和教学班类似于“国贸[1901-1903]班,国贸1901（香涛）班”的这种较为复杂的教学班字段是匹配不到的，但是数量极少，大约只有几个，可以忽略不计
|
||||
// 其实主要是教学班和课程名称和在一起了，不好分离，这里解析出来的字段有些是不怎么准确的，推荐使用课程课表来查询（但是得到的课貌似比这里的少（可能是线上课？））
|
||||
private static final Pattern pattern = Pattern.compile("(?<name>.*?[^ ]) ?(?<teachClass>教学班\\d+|\\d+班|临班\\d+|\\[\\d+-\\d+]班) (?<teacher>.*?) \\((?<weekString>.*?)周\\) ?(?<building>.*)"); |
||||
|
||||
@Override |
||||
public List<Course> parse(String html) throws ParseException { |
||||
return super.parse(html, courseGirdsXPath, pattern); |
||||
} |
||||
} |
||||
Loading…
Reference in new issue