Ubuntu 16.04 pandoc installation via cabal fails

Ubuntu 16.04 pandoc installation via cabal fails

[Uwe Brauer]

Hi

I need a decent version of cabal and I want to compile a filter and that
is why I cannot used stack as we discussed a week ago. So I finally
tried out the cabal way. Ubuntu ship the haskell platform with ghd
7.10.3 which should be ok,

However
cabal update
cabal install pandoc --enable-tests
gave the error
In-place registering pandoc-2.1.1...
Preprocessing executable 'pandoc' for pandoc-2.1.1...
[1 of 1] Compiling Main             ( pandoc.hs, dist/build/pandoc/pandoc-tmp/Main.o )
Linking dist/build/pandoc/pandoc ...
Preprocessing test suite 'test-pandoc' for pandoc-2.1.1...

test/Tests/Helpers.hs:16:8:
    Could not find module ‘Data.Algorithm.Diff’
    There are files missing in the ‘Diff-0.3.4@Diff_6besMvdnwF71rCmmNvH4et’ package,
    try running 'ghc-pkg check'.
    Use -v to see a list of the files searched for.
cabal: Error: some packages failed to install:
pandoc-2.1.1 failed during the building phase. The exception was:
ExitFailure 1

While
hc-pkg check
There are problems in package Diff-0.3.4:
  Warning: library-dirs: /home/oub/.cabal/lib/x86_64-linux-ghc-7.10.3/Diff-0.3.4-6besMvdnwF71rCmmNvH4et doesn't exist or isn't a directory
  import-dirs: /home/oub/.cabal/lib/x86_64-linux-ghc-7.10.3/Diff-0.3.4-6besMvdnwF71rCmmNvH4et doesn't exist or isn't a directory
  cannot find any of ["Data/Algorithm/Diff.hi","Data/Algorithm/Diff.p_hi","Data/Algorithm/Diff.dyn_hi"]
  cannot find any of ["Data/Algorithm/DiffOutput.hi","Data/Algorithm/DiffOutput.p_hi","Data/Algorithm/DiffOutput.dyn_hi"]
  cannot find any of ["Data/Algorithm/DiffContext.hi","Data/Algorithm/DiffContext.p_hi","Data/Algorithm/DiffContext.dyn_hi"]
  cannot find any of ["libHSDiff-0.3.4-6besMvdnwF71rCmmNvH4et.a","libHSDiff-0.3.4-6besMvdnwF71rCmmNvH4et.p_a","libHSDiff-0.3.4-6besMvdnwF71rCmmNvH4et-ghc7.10.3.so","libHSDiff-0.3.4-6besMvdnwF71rCmmNvH4et-ghc7.10.3.dylib","HSDiff-0.3.4-6besMvdnwF71rCmmNvH4et-ghc7.10.3.dll"] on library path

The following packages are broken, either because they have a problem
listed above, or because they depend on a broken package.
Diff-0.3.4


So I am stack at this point. Any ideas?

Uwe Brauer


-- 
You received this message because you are subscribed to the Google Groups "pandoc-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pandoc-discuss+unsubscribe-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
To post to this group, send email to pandoc-discuss-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
To view this discussion on the web visit https://groups.google.com/d/msgid/pandoc-discuss/87inb767g7.fsf%40mat.ucm.es.
For more options, visit https://groups.google.com/d/optout.

Re: Ubuntu 16.04 pandoc installation via cabal fails

[John MACFARLANE]

Try removing/backing up your ~/.cabal and ~/.ghc directories
(if present) and trying again.


+++ Uwe Brauer [Feb 08 18 12:34 ]:
>
>
>Hi
>
>I need a decent version of cabal and I want to compile a filter and that
>is why I cannot used stack as we discussed a week ago. So I finally
>tried out the cabal way. Ubuntu ship the haskell platform with ghd
>7.10.3 which should be ok,
>
>However
>cabal update
>cabal install pandoc --enable-tests
>gave the error
>In-place registering pandoc-2.1.1...
>Preprocessing executable 'pandoc' for pandoc-2.1.1...
>[1 of 1] Compiling Main             ( pandoc.hs, dist/build/pandoc/pandoc-tmp/Main.o )
>Linking dist/build/pandoc/pandoc ...
>Preprocessing test suite 'test-pandoc' for pandoc-2.1.1...
>
>test/Tests/Helpers.hs:16:8:
>    Could not find module ‘Data.Algorithm.Diff’
>    There are files missing in the ‘Diff-0.3.4@Diff_6besMvdnwF71rCmmNvH4et’ package,
>    try running 'ghc-pkg check'.
>    Use -v to see a list of the files searched for.
>cabal: Error: some packages failed to install:
>pandoc-2.1.1 failed during the building phase. The exception was:
>ExitFailure 1
>
>While
>hc-pkg check
>There are problems in package Diff-0.3.4:
>  Warning: library-dirs: /home/oub/.cabal/lib/x86_64-linux-ghc-7.10.3/Diff-0.3.4-6besMvdnwF71rCmmNvH4et doesn't exist or isn't a directory
>  import-dirs: /home/oub/.cabal/lib/x86_64-linux-ghc-7.10.3/Diff-0.3.4-6besMvdnwF71rCmmNvH4et doesn't exist or isn't a directory
>  cannot find any of ["Data/Algorithm/Diff.hi","Data/Algorithm/Diff.p_hi","Data/Algorithm/Diff.dyn_hi"]
>  cannot find any of ["Data/Algorithm/DiffOutput.hi","Data/Algorithm/DiffOutput.p_hi","Data/Algorithm/DiffOutput.dyn_hi"]
>  cannot find any of ["Data/Algorithm/DiffContext.hi","Data/Algorithm/DiffContext.p_hi","Data/Algorithm/DiffContext.dyn_hi"]
>  cannot find any of ["libHSDiff-0.3.4-6besMvdnwF71rCmmNvH4et.a","libHSDiff-0.3.4-6besMvdnwF71rCmmNvH4et.p_a","libHSDiff-0.3.4-6besMvdnwF71rCmmNvH4et-ghc7.10.3.so","libHSDiff-0.3.4-6besMvdnwF71rCmmNvH4et-ghc7.10.3.dylib","HSDiff-0.3.4-6besMvdnwF71rCmmNvH4et-ghc7.10.3.dll"] on library path
>
>The following packages are broken, either because they have a problem
>listed above, or because they depend on a broken package.
>Diff-0.3.4
>
>
>So I am stack at this point. Any ideas?
>
>Uwe Brauer
>
>
>-- 
>You received this message because you are subscribed to the Google Groups "pandoc-discuss" group.
>To unsubscribe from this group and stop receiving emails from it, send an email to pandoc-discuss+unsubscribe-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
>To post to this group, send email to pandoc-discuss-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
>To view this discussion on the web visit https://groups.google.com/d/msgid/pandoc-discuss/87inb767g7.fsf%40mat.ucm.es.
>For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups "pandoc-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pandoc-discuss+unsubscribe-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
To post to this group, send email to pandoc-discuss-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
To view this discussion on the web visit https://groups.google.com/d/msgid/pandoc-discuss/20180208211548.GE1526%40protagoras.
For more options, visit https://groups.google.com/d/optout.

[compiled filter does not work anymore] (was: Ubuntu 16.04 pandoc installation via cabal fails)

[Uwe Brauer]

   > Try removing/backing up your ~/.cabal and ~/.ghc directories
   > (if present) and trying again.

That seems to work, thanks very much.

Now  I have pandoc 2.1.1 installed, however the
filter
,----
| 
| import Text.Pandoc.JSON
|     
| main = toJSONFilter stripmath
|     where stripmath (Math _ _) = Str "FORMULA"
|           stripmath x = x
`----

Which I compiled via
ghc --make stripmath.hs

Results in an excutable which I use as

pandoc test.tex  -t json | /home/oub/ALLES/HGs/scripts/stripmath64bit | pandoc -f json -t plain -o test.txt

But I receive now the error:


Error at "source" (line 96, column 46):
unexpected %
\cite{oliynyk16:_futur_flrw,oliynyk15:_newton%
                                             ^
stripmath64bit: Error in $: not enough input
JSON parse error


The relevant code looks like

\cite{oliynyk16:_futur_flrw,oliynyk15:_newton%
}


But I can run this file with that bibtex entry perfectly with my old
(32bit compiled against 1.17 pandoc version)

So what shall I do?

Thanks

Uwe Brauer

Re: (was: Ubuntu 16.04 pandoc installation via cabal fails)

[John MacFarlane]

The filter has to be compiled againts the same version of
pandoc-types as the binary you're using.  I'm guessing
this hasn't happened in this case, but I'd need to
know more about your package versions.

One solution would be to remove your pandoc 2.1.1
(assuming you installed from our binary package)
and use cabal to install pandoc itself.

+++ Uwe Brauer [Feb 10 18 11:58 ]:
>
>   > Try removing/backing up your ~/.cabal and ~/.ghc directories
>   > (if present) and trying again.
>
>That seems to work, thanks very much.
>
>Now  I have pandoc 2.1.1 installed, however the
>filter
>,----
>|
>| import Text.Pandoc.JSON
>|
>| main = toJSONFilter stripmath
>|     where stripmath (Math _ _) = Str "FORMULA"
>|           stripmath x = x
>`----
>
>Which I compiled via
>ghc --make stripmath.hs
>
>Results in an excutable which I use as
>
>pandoc test.tex  -t json | /home/oub/ALLES/HGs/scripts/stripmath64bit | pandoc -f json -t plain -o test.txt
>
>But I receive now the error:
>
>
>Error at "source" (line 96, column 46):
>unexpected %
>\cite{oliynyk16:_futur_flrw,oliynyk15:_newton%
>                                             ^
>stripmath64bit: Error in $: not enough input
>JSON parse error
>
>
>The relevant code looks like
>
>\cite{oliynyk16:_futur_flrw,oliynyk15:_newton%
>}
>
>
>But I can run this file with that bibtex entry perfectly with my old
>(32bit compiled against 1.17 pandoc version)
>
>So what shall I do?
>
>Thanks
>
>Uwe Brauer
>
>-- 
>You received this message because you are subscribed to the Google Groups "pandoc-discuss" group.
>To unsubscribe from this group and stop receiving emails from it, send an email to pandoc-discuss+unsubscribe-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
>To post to this group, send email to pandoc-discuss-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
>To view this discussion on the web visit https://groups.google.com/d/msgid/pandoc-discuss/87eflt1574.fsf_-_%40mat.ucm.es.
>For more options, visit https://groups.google.com/d/optout.

Re:

[Uwe Brauer]

>>> "John" == John MacFarlane <jgm-TVLZxgkOlNX2fBVCVOL8/A@public.gmane.org> writes:

   > The filter has to be compiled againts the same version of
   > pandoc-types as the binary you're using.  I'm guessing
   > this hasn't happened in this case, but I'd need to
   > know more about your package versions.

   > One solution would be to remove your pandoc 2.1.1
   > (assuming you installed from our binary package)
   > and use cabal to install pandoc itself.

I desinstalled all the deb binary pandoc packages deleted .cabal and
.ghc
and run

cabal update
cabal install pandoc --enable-tests
I compile the filter,

ghc --make stripmath.hs

This is the filter which worked fine in pandoc 1.17
,----
| 
| import Text.Pandoc.JSON
|     
| main = toJSONFilter stripmath
|     where stripmath (Math _ _) = Str "FORMULA"
|           stripmath x = x
| 
`----

I run it
pandoc test.tex  -t json | /home/oub/ALLES/HGs/scripts/stripmath | pandoc -f json -t plain -o test.txt'


and again the error message pops up



Error at "source" (line 96, column 46):
unexpected %
\cite{oliynyk16:_futur_flrw,oliynyk15:_newton%
                                             ^
stripmath64bit: Error in $: not enough input
JSON parse error


Again everything was fine in 1.17

Uwe Brauer 

Re:

[John MacFarlane]

It looks like this error may have something to do with
the specifics of your filter; perhaps you could post it?

+++ Uwe Brauer [Feb 13 18 12:07 ]:
>>>> "John" == John MacFarlane <jgm-TVLZxgkOlNX2fBVCVOL8/A@public.gmane.org> writes:
>
>   > The filter has to be compiled againts the same version of
>   > pandoc-types as the binary you're using.  I'm guessing
>   > this hasn't happened in this case, but I'd need to
>   > know more about your package versions.
>
>   > One solution would be to remove your pandoc 2.1.1
>   > (assuming you installed from our binary package)
>   > and use cabal to install pandoc itself.
>
>I desinstalled all the deb binary pandoc packages deleted .cabal and
>.ghc
>and run
>
>cabal update
>cabal install pandoc --enable-tests
>I compile the filter,
>
>ghc --make stripmath.hs
>
>This is the filter which worked fine in pandoc 1.17
>,----
>|
>| import Text.Pandoc.JSON
>|
>| main = toJSONFilter stripmath
>|     where stripmath (Math _ _) = Str "FORMULA"
>|           stripmath x = x
>|
>`----
>
>I run it
>pandoc test.tex  -t json | /home/oub/ALLES/HGs/scripts/stripmath | pandoc -f json -t plain -o test.txt'
>
>
>and again the error message pops up
>
>
>
>Error at "source" (line 96, column 46):
>unexpected %
>\cite{oliynyk16:_futur_flrw,oliynyk15:_newton%
>                                             ^
>stripmath64bit: Error in $: not enough input
>JSON parse error
>
>
>Again everything was fine in 1.17
>
>Uwe Brauer
>
>-- 
>You received this message because you are subscribed to the Google Groups "pandoc-discuss" group.
>To unsubscribe from this group and stop receiving emails from it, send an email to pandoc-discuss+unsubscribe-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
>To post to this group, send email to pandoc-discuss-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
>To view this discussion on the web visit https://groups.google.com/d/msgid/pandoc-discuss/873725jggo.fsf%40mat.ucm.es.
>For more options, visit https://groups.google.com/d/optout.

Re:

[Uwe Brauer]

>>> "John" == John MacFarlane <jgm-TVLZxgkOlNX2fBVCVOL8/A@public.gmane.org> writes:

    > It looks like this error may have something to do with
    > the specifics of your filter; perhaps you could post it?

But I did it my last post. Ok here it comes again

,----
| 
| import Text.Pandoc.JSON
|     
| main = toJSONFilter stripmath
|     where stripmath (Math _ _) = Str "FORMULA"
|           stripmath x = x
`----


You wrote that filter a couple of months ago or maybe more than  a year
ago.

Re:

[Jesse Rosenthal]

Uwe Brauer <oub-YB6e1s5WF/He5aOfsHch1g@public.gmane.org> writes:

>>>> "John" == John MacFarlane <jgm-TVLZxgkOlNX2fBVCVOL8/A@public.gmane.org> writes:

> Error at "source" (line 96, column 46):
> unexpected %
> \cite{oliynyk16:_futur_flrw,oliynyk15:_newton%

The error here seems to be not in the filter, but in the fact that
pandoc can't parse the LaTeX. Note that running `pandoc -f latex` on
this input:

    \cite{oliynyk16:_futur_flrw,oliynyk15:_newton%
    }

will produce the above error. And then since you have no output, you're
passing nothing through the pipe to your script. Try running just pandoc
on the file, and see if this is the issue.

Re:

[Jesse Rosenthal]

Uwe Brauer <oub-YB6e1s5WF/He5aOfsHch1g@public.gmane.org> writes:
> but why does it work with the filter compiled against pandoc 1.17?

Again, I don't think it has anything to do with the filter. I assume the
LaTeX reader must have changed. Did you try running the file through a
recent pandoc without the filter?

(The JSON format changed in 1.18, so I'm almost certain you couldn't be
running the output of a new pandoc through a filter compiled against
1.17.)

Re:

[Uwe Brauer]

>>> "Jesse" == Jesse Rosenthal <jrosenthal-4GNroTWusrE@public.gmane.org> writes:

   > Uwe Brauer <oub-YB6e1s5WF/He5aOfsHch1g@public.gmane.org> writes:
   >>>>> "John" == John MacFarlane <jgm-TVLZxgkOlNX2fBVCVOL8/A@public.gmane.org> writes:

   >> Error at "source" (line 96, column 46):
   >> unexpected %
   >> \cite{oliynyk16:_futur_flrw,oliynyk15:_newton%

   > The error here seems to be not in the filter, but in the fact that
   > pandoc can't parse the LaTeX. Note that running `pandoc -f latex` on
   > this input:

   >     \cite{oliynyk16:_futur_flrw,oliynyk15:_newton%
   >     }

   > will produce the above error. And then since you have no output,
   > you're passing nothing through the pipe to your script. Try running
   > just pandoc on the file, and see if this is the issue.

Ok now I understand, I am sorry I have not read your email carefully
enough before.

I have the different filters in different machine, I will now run the
old filter and see tomorrow whether the new reader can deal with it.

In case it can't what would be the solution. I was so happy with the old
filter and reader because it solved an important issue for me.

Uwe 

Re:

[Uwe Brauer]

>>> "Jesse" == Jesse Rosenthal <jrosenthal-4GNroTWusrE@public.gmane.org> writes:

   > Uwe Brauer <oub-YB6e1s5WF/He5aOfsHch1g@public.gmane.org> writes:
   >> but why does it work with the filter compiled against pandoc 1.17?

   > Again, I don't think it has anything to do with the filter. I assume the
   > LaTeX reader must have changed. Did you try running the file through a
   > recent pandoc without the filter?

   > (The JSON format changed in 1.18, so I'm almost certain you couldn't be
   > running the output of a new pandoc through a filter compiled against
   > 1.17.)

Maybe I misunderstood you I wanted to combine the old filter with the
new reader.

Old filter and old reader are involved via

pandoc Report.tex -t json /home/oub/ALLES/HGs/scripts/stripmath32bit  |
pandoc -f json -t plain -o out.txt

So I tried 
pandoc Report.tex -t json /home/oub/ALLES/HGs/scripts/stripmath32bit

But I got 
pandoc: Cannot decode byte '\xc7': Data.Text.Internal.Encoding.Fusion.streamUtf8: Invalid UTF-8 stream

Now I am even more puzzled.

[Bug/Error in latex reader, but syntax ok] (was: )

[Uwe Brauer]

    > It looks like this error may have something to do with
    > the specifics of your filter; perhaps you could post it?

I followed Jesse's advice and indeed there is a problem with the latex
reader in new versions of pandoc. Syntactical correct latex,

Such as

\cite{oliynyk16:_futur_flrw,oliynyk15:_newton%
}


So running

Pandoc test.tex -o test.txt
with such a file leads to

pandoc Report.tex -o test.txt

Error at "source" (line 96, column 46):
unexpected %
\cite{oliynyk16:_futur_flrw,oliynyk15:_newton%


This is a but in pandoc, I think.

Re: (was: )

[John MacFarlane]

Looks like a bug, yes.  Can you report it on the github tracker?

+++ Uwe Brauer [Feb 15 18 09:53 ]:
>
>    > It looks like this error may have something to do with
>    > the specifics of your filter; perhaps you could post it?
>
>I followed Jesse's advice and indeed there is a problem with the latex
>reader in new versions of pandoc. Syntactical correct latex,
>
>Such as
>
>\cite{oliynyk16:_futur_flrw,oliynyk15:_newton%
>}
>
>
>So running
>
>Pandoc test.tex -o test.txt
>with such a file leads to
>
>pandoc Report.tex -o test.txt
>
>Error at "source" (line 96, column 46):
>unexpected %
>\cite{oliynyk16:_futur_flrw,oliynyk15:_newton%
>
>
>This is a but in pandoc, I think.
>
>
>-- 
>You received this message because you are subscribed to the Google Groups "pandoc-discuss" group.
>To unsubscribe from this group and stop receiving emails from it, send an email to pandoc-discuss+unsubscribe-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
>To post to this group, send email to pandoc-discuss-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
>To view this discussion on the web visit https://groups.google.com/d/msgid/pandoc-discuss/87inay7hx0.fsf_-_%40mat.ucm.es.
>For more options, visit https://groups.google.com/d/optout.

Re:

[Uwe Brauer]

>>> "John" == John MacFarlane <jgm-TVLZxgkOlNX2fBVCVOL8/A@public.gmane.org> writes:

   > Looks like a bug, yes.  Can you report it on the github tracker?

Done!