Incorrect parsing of raw multiline strings

[ayr0]

IPython incorrectly parses the following input

test = r'''%@article {Cat61,
%    AUTHOR = {Cattabriga, Lamberto},
'''

I am parsing LaTeX, and IPython grabs each line starting with '%' as an ipython_magic function instead of correctly recognizing that '%' appears in a raw string.

"%    AUTHOR = {Cattabriga, Lamberto},\nget_ipython().magic(u'TITLE = {Su un problema al contorno relativo al sistema di equazioni di')\n"

The bad parsing is triggered by the comma and newline at the end of the first line.

test = r'''%test,
%test2
'''
"%test,\nget_ipython().magic(u'test2')\n"

test = r'''%test
%test2
'''
'%test\n%test2\n'

[ivanov]

@ayr0 thanks for the report, but can you please provide more detail about what version of IPython you are running (ipython --version), and what client you are using (good ole terminal ipython, ipython qtconsole, console, web notebook?)

I am unable to reproduce the error you see on any of the recent version (git master, 0.13, 0.12, or 0.11)

@takluyver recently did some work on how these magics are processed, so he might have an idea.

[jasongrout]

Hi Ryan :)

I can confirm that in IPython notebook, latest master (commit 6c36cd9), I have the following input/output in the notebook:

test = r'''%@article {Cat61,
%    AUTHOR = {Cattabriga, Lamberto},
'''

Then test is "%@article {Cat61,\nget_ipython().magic(u'AUTHOR = {Cattabriga, Lamberto},')\n" (note the magic call...)

If input is:

test = r'''%test,
%test2
'''

then test is "%test,\nget_ipython().magic(u'test2')\n"

I think this issue is probably related to gh-2628, and is probably a result of the code at https://github.com/ipython/ipython/blob/master/IPython/core/inputsplitter.py#L908 (just like gh-2628 is triggered by an omission at that line, as you can see from gh-2428).

[takluyver]

This is essentially the same issue as #2612, so I'm going to close this, and we'll track it there. I have some ideas about how to fix it, but I want to get PR #2447 sorted out first.