Extract abstract from plain ascii draft

[ietf-svn-bot]

owner:jtarragon@yaco.es resolution_fixed type_enhancement | by esanchez@yaco.es

---

We have to extract the abstract of a draft. You can view how the old perl tool does the work at http://wiki.tools.ietf.org/tools/ietfdb/browser/branch/legacy/idst/validate.cgi#L254

You have ready a parse_normal_005_abstract method in plain_parser to do the job. Please take into account the following facts:

• You have available self.lines and self.full_text with the draft splited into lines and as a whole
• You probably have some aid functions inside parse_normal_004_authors that you could refactor as a plain_parser methods to help you parsing the abstract (like the one that strip out the page headers). Also the authors parser has some pyflakes errors, please, review it.

---

Issue migrated from trac:587 at 2022-03-04 01:43:35 +0000

[ietf-svn-bot]

@henrik@levkowetz.com commented

---

Please don't rely too much on the Perl code. After a number of tweaks over the last couple of years it doesn't break quite as much as it used to, but it's still not necessarily the best approach.

Anyway, whatever approach you take, please test the resulting code against
1000 or so draft instances from the last 2 years. If it does a good job on
those, we're probably OK.

[ietf-svn-bot]

@esanchez@yaco.es commented

---

Please, forget the parse_normal_005_abstract_method. Implement the method inside the Draft class in utils/draft.py

Take a look to ietf.idtracker.models.InternetDraft.clean_abstract method. Add this logic to the method that retrieves the abstract so it needs no other process later.

[ietf-svn-bot]

@jtarragon@yaco.es commented

---

From ffbea31:

Implement abstract parsing in draft utils. See #587.

[ietf-svn-bot]

@esanchez@yaco.es changed status from new to closed

[ietf-svn-bot]

@esanchez@yaco.es set resolution to fixed

[ietf-svn-bot]

@esanchez@yaco.es commented

---

Fixed in 90e349c:

Draft class already does a good parsing and cleaning of the text so we can retrieve the abstract in a KISS way. Fixes #587

[ietf-svn-bot]

@esanchez@yaco.es commented

---

I tested the new method over the 1000 most recent drafts (since 2010-10-25, I'm a bit outdated :P) and it seems to work properly.

• No draft return a blank Abstract
• One draft return an abstract longer that 20 lines (but I check it manually and it actually has an abstract of 29 lines)
• I print the 100 first abstracts and they seem to be ok

[ietf-svn-bot]

@esanchez@yaco.es changed status from closed to reopened

[ietf-svn-bot]

@esanchez@yaco.es changed resolution from fixed to ``

[ietf-svn-bot]

@esanchez@yaco.es commented

---

I see that the abstract includes the first blank line(s). This could be improved.

[ietf-svn-bot]

@esanchez@yaco.es changed status from reopened to closed

[ietf-svn-bot]

@esanchez@yaco.es changed resolution from `` to fixed

[ietf-svn-bot]

@esanchez@yaco.es commented

---

Fixed in 9fe6736:

Trhow away the first blank lines of an abstract. Fixes #587

[ietf-svn-bot]

@esanchez@yaco.es changed status from closed to reopened

[ietf-svn-bot]

@esanchez@yaco.es changed resolution from fixed to ``