Interfacing pyamihtml to docanalysis

[petermr]

pyamihtml by default creates an html file with heuristic sections, often based on decimals. docanalysis takes xml sections based on PMC/JATS sectioning. These sections are listed in docanalysis/glob_trail.py , e.g.:

# define constants
ABS = ['*abstract.xml']
ACK = ['*ack.xml']
AFF = ['*aff.xml']
AUT = ['*contrib-group.xml']
CON = ['*conclusion*/*.xml']
DIS = ['*discussion*/**/*_title.xml', '*discussion*/**/*_p.xml'] # might bring unwanted sections like tables, fig. captions etc. Maybe get only title and paragraphs?
ETH = ['*ethic*/*.xml']
FIG = ['*fig*.xml']
INT = ['*introduction*/*.xml', '*background*/*.xml']
KEY = ['*kwd-group.xml']
MET = ['*method*/*.xml', '*material*/*.xml'] # also gets us supplementary material. Not sure how to exclude them
RES = ['*result*/*/*_title.xml', '*result*/*/*_p.xml'] # not sure if we should use recursive globbing or not. 
TAB = ['*table*.xml']
TIL = ['*article-meta/*title-group.xml']

proposal 1

Define these sections in a JSON file read in by docanalysis so that

• the regexes can be edited without affecting the code
• new keywords can be added

proposal 2

• make pyamihtml output a directory structure compatible with JATS/docanalysis
  This requires discrete sections in a hierarchy which are labelled *.xml.
• make these sections as well as the normal HTML.