Fixes parser for issue HTML

This also adds .journal_title as an attribute to the Issue object
This commit is contained in:
Nemo 2020-06-30 15:19:12 +05:30
parent 870ed3080d
commit 919c8ac43f
5 changed files with 1361 additions and 21 deletions

1263
spec/fixtures/issue-35852.html vendored Normal file

File diff suppressed because it is too large Load Diff

View File

@ -37,4 +37,25 @@ describe Muse::Dl::Issue do
it "should parse publisher" do
issue.publisher.should eq "Johns Hopkins University Press"
end
it "should parse the journal title" do
issue.journal_title.should eq "portal: Libraries and the Academy"
end
it "should parse non-numbered issues" do
WebMock.stub(:get, "https://muse.jhu.edu/issue/35852")
.to_return(body: File.new("spec/fixtures/issue-35852.html").gets_to_end)
issue = Muse::Dl::Issue.new "35852"
issue.parse
issue.volume.should eq "1"
issue.number.should eq "2"
issue.date.should eq "2016"
issue.info["ISSN"].should eq "2474-9419"
issue.info["Print ISSN"].should eq "2474-9427"
issue.info["Launched on MUSE"].should eq "2017-02-21"
issue.info["Open Access"].should eq "Yes"
issue.title.should eq "Volume 1, Issue 2, 2016"
issue.journal_title.should eq "Constitutional Studies"
end
end

View File

@ -3,8 +3,8 @@ require "./issue.cr"
module Muse::Dl
class Article
@id : String
getter :id
getter id : String
setter title : String | Nil, start_page : Int32 | Nil, end_page : Int32 | Nil
def initialize(id : String)
@id = id

View File

@ -1,29 +1,26 @@
require "./thing.cr"
"./thing.cr"
require "./fetch.cr"
require "./article.cr"
module Muse::Dl
class Issue
@id : String
@title : String | Nil
@articles : Array(Muse::Dl::Article)
@url : String
@info : Hash(String, String)
@summary : String | Nil
@publisher : String | Nil
@volume : String | Nil
@number : String | Nil
@date : String | Nil
@issues : Array(Muse::Dl::Issue)
getter :id, :title, :articles, :url, :summary, :publisher, :info, :volume, :number, :date
getter id : String,
title : String | Nil,
articles : Array(Muse::Dl::Article),
url : String,
summary : String | Nil,
publisher : String | Nil,
info : Hash(String, String),
volume : String | Nil,
number : String | Nil,
date : String | Nil,
journal_title : String | Nil
def initialize(id : String)
@id = id
@url = "https://muse.jhu.edu/issue/#{id}"
@info = Hash(String, String).new
@articles = [] of Muse::Dl::Article
@issues = [] of Muse::Dl::Issue
end
def parse
@ -42,17 +39,38 @@ module Muse::Dl
unless t.nil?
@volume = /Volume (\d+)/.match(t).try &.[1]
@number = /Number (\d+)/.match(t).try &.[1]
@number = /Issue (\d+)/.match(t).try &.[1] unless @number
@date = /((January|February|March|April|May|June|July|August|September|October|November|December) (\d+))/.match(t).try &.[1]
@date = /(\d{4})/.match(t).try &.[1] unless @date
end
end
def parse_contents(myhtml : Myhtml::Parser)
myhtml.css("#available_issues_list_text a").each do |a|
link = a.attribute_by("href").to_s
journal_title_a = myhtml.css("#journal_banner_title a").first
if journal_title_a
@journal_title = journal_title_a.inner_text
end
myhtml.css(".articles_list_text ol").each do |ol|
link = ol.css("li.title a").first
title = link.inner_text
matches = /\/issue\/(\d+)/.match link
pages = ol.css("li.pg").first.try &.inner_text
matches = /(\d+)-(\d+)/.match pages
if matches
@issues.push Muse::Dl::Issue.new matches[1]
start_page = matches[1].to_i
end_page = matches[2].to_i
end
ol.css("a").each do |l|
url = l.attribute_by("href").to_s
matches = /\/article\/(\d+)\/pdf/.match url
if matches
a = Muse::Dl::Article.new matches[1]
a.title = title
a.start_page = start_page if start_page
a.end_page = end_page if end_page
@articles.push a
end
end
end
end

View File

@ -71,6 +71,44 @@ module Muse::Dl
# Needed because of https://github.com/crystal-lang/crystal/issues/7777
FileUtils.cp source, destination
FileUtils.rm source if parser.cleanup
elsif thing.is_a? Muse::Dl::Issue
# Will have no effect if parser has a custom title
parser.output = Util.slug_filename "#{thing.title}.pdf"
# If file exists and we can't clobber
if File.exists?(parser.output) && parser.clobber == false
STDERR.puts "Skipping #{url}, File already exists: #{parser.output}"
return
end
temp_stitched_file = nil
pdf_builder = Pdftk.new(parser.tmp)
# ## TODO till 111
thing.issues.each do |issue|
begin
Fetch.save_issue(parser.tmp, chapter[0], chapter[1], parser.cookie, parser.bookmarks, parser.strip_first)
rescue e : Muse::Dl::Errors::MuseCorruptPDF
STDERR.puts "Got a 'Unable to construct chapter PDF' error from MUSE, skipping: #{url}"
return
end
end
chapter_ids = thing.chapters.map { |c| c[0] }
# Stitch the PDFs together
temp_stitched_file = pdf_builder.stitch chapter_ids
pdf_builder.add_metadata(temp_stitched_file, parser.output, thing)
temp_stitched_file.delete if temp_stitched_file
puts "--dont-strip-first-page was on. Please validate PDF file for any errors." if parser.strip_first
puts "DL: #{url}. Saved final output to #{parser.output}"
# Cleanup the chapter files
if parser.cleanup
thing.chapters.each do |c|
Fetch.cleanup(parser.tmp, c[0])
end
end
####
end
end