Marty
/
pubs-extract
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases 2 Activity

Delineate internal from interface methods

This commit is contained in:
Marty Oehme 2022-12-22 18:57:41 +01:00
parent 2b4cec9c88
commit 4f67c13cc0
Signed by: Marty
GPG Key ID: 73BA40D5AFAF49C9
1 changed files with 15 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

32
extract/extract.py
View File

@ -1,7 +1,5 @@
import os import os
import sys
import argparse import argparse
from typing import Dict, List, Tuple
# from subprocess import Popen, PIPE, STDOUT # from subprocess import Popen, PIPE, STDOUT
# from pipes import quote as shell_quote # from pipes import quote as shell_quote
@ -9,12 +7,12 @@ from typing import Dict, List, Tuple
import fitz import fitz
# from ... import uis # from ... import uis
from ...plugins import PapersPlugin from pubs.plugins import PapersPlugin
from ...events import PaperChangeEvent, PostCommandEvent from pubs.events import PaperChangeEvent, PostCommandEvent
from ... import repo from pubs import repo
from ...utils import resolve_citekey_list from pubs.utils import resolve_citekey_list
from ...content import write_file from pubs.content import write_file
@ -77,37 +75,37 @@ class ExtractPlugin(PapersPlugin):
) )
if not citekeys: if not citekeys:
return return
papers = self.gather_papers(citekeys) papers = self._gather_papers(citekeys)
all_annotations = self.extract(papers) all_annotations = self.extract(papers)
if args.write: if args.write:
self.to_notes(conf, all_annotations, args.edit) self._to_notes(conf, all_annotations, args.edit)
else: else:
self.to_stdout(all_annotations) self._to_stdout(all_annotations)
self.repository.close() self.repository.close()
def extract(self, papers): def extract(self, papers):
papers_annotated = [] papers_annotated = []
for paper in papers: for paper in papers:
file = self.get_file(paper) file = self._get_file(paper)
try: try:
papers_annotated.append((paper, self.get_annotations(file))) papers_annotated.append((paper, self._get_annotations(file)))
except fitz.FileDataError as e: except fitz.FileDataError as e:
print(f"ERROR: Document {file} is broken: {e}") print(f"ERROR: Document {file} is broken: {e}")
return papers_annotated return papers_annotated
def gather_papers(self, citekeys): def _gather_papers(self, citekeys):
papers = [] papers = []
for key in citekeys: for key in citekeys:
papers.append(self.repository.pull_paper(key)) papers.append(self.repository.pull_paper(key))
return papers return papers
def get_file(self, paper): def _get_file(self, paper):
path = self.broker.real_docpath(paper.docpath) path = self.broker.real_docpath(paper.docpath)
if not path: if not path:
self.ui.error(f"{paper.citekey} has no valid document.") self.ui.error(f"{paper.citekey} has no valid document.")
return path return path
def get_annotations(self, filename): def _get_annotations(self, filename):
annotations = [] annotations = []
with fitz.Document(filename) as doc: with fitz.Document(filename) as doc:
for page in doc: for page in doc:
@ -119,7 +117,7 @@ class ExtractPlugin(PapersPlugin):
annotations.append(f"[{page.number}] {content}") annotations.append(f"[{page.number}] {content}")
return annotations return annotations
def to_stdout(self, annotated_papers): def _to_stdout(self, annotated_papers):
for contents in annotated_papers: for contents in annotated_papers:
paper = contents[0] paper = contents[0]
annotations = contents[1] annotations = contents[1]
@ -129,7 +127,7 @@ class ExtractPlugin(PapersPlugin):
print(f'> "{annot}"') print(f'> "{annot}"')
print("") print("")
def to_notes(self, conf, annotated_papers, edit=False): def _to_notes(self, conf, annotated_papers, edit=False):
for contents in annotated_papers: for contents in annotated_papers:
paper = contents[0] paper = contents[0]
annotations = contents[1] annotations = contents[1]