Commit 462f074a authored by Alexander Ross Melnick's avatar Alexander Ross Melnick
Browse files

LOTS of English data crawled

parent e4d135c1

.gitignore

+1 −2
Original line number Diff line number Diff line
@@ -7,8 +7,7 @@ CheckerCorrector/*.txt 
CheckerCorrector/*db

CheckerCorrector/corrector

CheckerCorrector/checker

Crawler/target/classes/ScratchCrawler.class

Crawler/target/classes/crawledData.txt

Crawler/target/*

*.db

*.java

*.class

Crawler/src/main/java/Debug.class

(352 B)

File changed.

No diff preview for this file type.

View original file
View changed file

Crawler/src/main/java/Makefile

+9 −7
Original line number Diff line number Diff line
@@ -2,20 +2,22 @@ JAVAC=javac 
JFLAGS=-Xlint:none -Xlint:-deprecation

LIBS=tika-app-2.9.2.jar



default: ScratchCrawler.class Debug.class RobotsTXT.class RegexParser.class

default: all



ScratchCrawler.class: ScratchCrawler.java

	$(JAVAC) $(JFLAGS) -cp $(LIBS) ScratchCrawler.java

all: Debug.class RobotsTXT.class RegexParser.class ScratchCrawler.class



ScratchCrawler.class: ScratchCrawler.java RobotsTXT.class Debug.class RegexParser.class

	$(JAVAC) $(JFLAGS) -cp .:$(LIBS) ScratchCrawler.java



Debug.class: Debug.java

	$(JAVAC) $(JFLAGS) -cp $(LIBS) Debug.java

	$(JAVAC) $(JFLAGS) -cp .:$(LIBS) Debug.java



RobotsTXT.class: RobotsTXT.java

	$(JAVAC) $(JFLAGS) -cp $(LIBS) RobotsTXT.java

	$(JAVAC) $(JFLAGS) -cp .:$(LIBS) RobotsTXT.java



RegexParser.class: RegexParser.java

	$(JAVAC) $(JFLAGS) -cp $(LIBS) RegexParser.java

	$(JAVAC) $(JFLAGS) -cp .:$(LIBS) RegexParser.java



clean:

	rm -f *.class

	del *.class

	rm -f *.class

Crawler/src/main/java/RegexParser.class

−1 B (3.84 KiB)

File changed.

No diff preview for this file type.

View original file
View changed file

Crawler/src/main/java/RobotsTXT.class

(1.19 KiB)

File changed.

No diff preview for this file type.

View original file
View changed file