s470623/twilight-library
1
0
Fork 0
forked from filipg/twilight-library
Code Issues Pull Requests Projects Releases Wiki Activity

title extraction

Browse Source
This commit is contained in:
s470623 2022-06-15 22:27:32 +02:00
parent 1f23d94e7c
commit 0114a102d4
1 changed files with 3 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
app/teatrLalek.hs
View File

@ -17,7 +17,7 @@ extractRecords = extractLinks "//a[contains(@href, '.pdf')]"
toShadowItem :: String -> ShadowItem toShadowItem :: String -> ShadowItem
toShadowItem url = toShadowItem url =
(defaultShadowItem url []) { (defaultShadowItem url title) {
originalDate = Just date, originalDate = Just date,
itype = "periodical", itype = "periodical",
format = Just "pdf", format = Just "pdf",
@ -25,6 +25,8 @@ toShadowItem url =
} }
where where
date = last $ getAllTextMatches $ url =~ "(19[0-9][0-9]|20[0-9][0-9])" :: String date = last $ getAllTextMatches $ url =~ "(19[0-9][0-9]|20[0-9][0-9])" :: String
titleToProcess = last $ getAllTextMatches $ url =~ "/[a-zA-Z ]+[-_]" :: String
title = titleToProcess =~ "[a-zA-Z ]+" :: String
main = do main = do