Adding info sections does not seem to work

[BlackEdder]

I have the feeling it is looking for:

header:
  -section 

While info sections probably don't start with a hyphen.

[tobimensch]

Hello,

well, it's indeed looking for something roughly like you suggest.

Could you provide an example of "--help" output you want to parse and tell me what part of it should be considered "info section".

Greetings

[BlackEdder]

I wanted to use it on the help message for (see below): https://github.com/BlackEdder/plotd

And I basically wanted everything after "Data format:" into its own section. I tried renaming Data format to examples, but that also didn't work.

Usage: plotcli [-f] [-o OUTPUT] [-d FORMAT] [-b BOUNDS] [--xlabel XLABEL] [--ylabel YLABEL] [--margin-bounds MARGINBOUNDS] [--image IMAGEFORMAT]

Plotcli is a plotting program that will plot data from provided data streams (files). It will ignore any lines it doesn't understand, making it possible to feed it "dirty" streams/files. All options can also be provided within the stream by using the prefix #plotcli (e.g. #plotcli -d x,y).

Options:
  -f          Follow the stream, i.e. keep listening for new lines.
  -d FORMAT   String describing the content of each row. Different row formats supported: x, y and h, with h indication histogram data. For more information see Data format section.
  -o OUTPUT       Outputfile (without extension).
  -b BOUNDS   Give specific bounds for the plot in a comma separated list (min_x,max_x,min_y,max_y).
  --xlabel XLABEL
  --ylabel YLABEL
  --margin-bounds MARGINBOUNDS  Specific bounds (in pixel size) for the margins. Format (all in pixels): xmargin,xwidth,ymargin,yheight. Default values 70,400,70,400.
  --image IMAGEFORMAT  Format of the resulting image (png/pdf/svg).

Data format:
  Using -d it is possible to specify what each column in your data file represents. Supported formats are:

  x,y   The x and y coordinate for points
  lx,ly Line data
  h     Histogram data
  hx,hy 2D Histogram data 
  ..      Extrapolate from previous options, i.e. x,y,.. -> x,y,x,y,..
  id    Default data id to use for this row of data (it is also possible to provide a column specific id (see Data ids below))
  pn    Default plot name to use for this row of data (it is also possible to provide a column specific id (see Plot ids below))

  Examples: x,y,y or h,x,y. When there are more ys provided than xs (or vice versa) the last x will be matched to all remaining ys.

  Data ids: plotcli by default does a good job of figuring out which x and y data belong together, but you can optionally provide an numeric id to make this completely clear. I.e. x1,y1. Data ids always need to directly follow the format type (before plot ids).

  Plot ids: if you want to plot the data to different figures you can add a letter/name at the end: xa,ya or x1a,y1a. This plot id will be appended to the OUTPUT file name. 

  Extrapolating (..): plotcli will try to extrapolate from your previous options. This also works for simple plot ids. I.e. if you want a separate histogram for each column: ha,hb,.. results in ha,hb,hc,hd,he etc. Other examples: y,.. -> y,y,y,y etc. x,y,y,.. -> x,y,y,y,y etc.

[tobimensch]

Thanks for your use case example. I'll try to make this work.

[tobimensch]

@BlackEdder

This issue should be fixed in master. I tested it with the help output you provided.

cli2man plotcli --info-section "Data format" should do the trick!

I know this is a bit late, but I guess better late then never right? :D

[tobimensch]

Closing this as this should be fixed now. But if you can't confirm it is fixed, please open another issue, or add a comment here.