Both report a list (sorted in descending order) of langauges used in a given dataset: GitHub reports the number of bytes whereas CLOC reports number of lines, but since we're interested in percentages, both are fine
The result is the generation of a new pie chart for each repository (plus a global one) with the language distribution.
When
Language statistics are always generated when SLOCs are counted with cloc, since in that case those statistics are basically free by further parsing CLOC's output.
If SLOCSs are counted with wc or not counted at all, the user can specify the new --lang command line argument to have the script fetch those informations from GitHub's APIs. Of course I also added a complementary --no-lang that skips language generation.
Note that since CLOC's output is much more detailed than GitHub's APIs, it always takes precedence: if CLOC is being used then the --lang and --no-lang options are ignored and if neither is provided, the prompt asking the user wether they want to generate language statistics is skipped.
Documentation
I updated the combined.svg used in the README and removed the other two unused SVGs
Also i changed the README to reflect the new command line arguments. I also made some somewhat random changes here and there, the bigger one being in the usage instructions. These changes have various reasons behind them but are mostly subjective and stylistic, so have a look at them and let me know if the changes are appreciated.
Output
I made some changes to the output the program produces:
I added an output section for the language statistics
Hid git clone output for cleaner interaction with the user
Added counters to graph generation so the user can see how it's coming along (it can take a while on slower machines)
Changed output so that each "section" (contributions, commits, SLOC, graphs, etc.) is separated by two newlines from the previous one and all tasks within a section (counting SLOC for a specific repo etc.) are indented
Changed report output such that global statistics are separated by per-repo statistics in each section by a newline, to make them stand out more
Language statistics
Generate language statistics for each repository and globally, to show how much of our code is made by each language, in percentage.
How
Both report a list (sorted in descending order) of langauges used in a given dataset: GitHub reports the number of bytes whereas CLOC reports number of lines, but since we're interested in percentages, both are fine
The result is the generation of a new pie chart for each repository (plus a global one) with the language distribution.
When
Language statistics are always generated when SLOCs are counted with
cloc, since in that case those statistics are basically free by further parsing CLOC's output.If SLOCSs are counted with
wcor not counted at all, the user can specify the new--langcommand line argument to have the script fetch those informations from GitHub's APIs. Of course I also added a complementary--no-langthat skips language generation.Note that since CLOC's output is much more detailed than GitHub's APIs, it always takes precedence: if CLOC is being used then the
--langand--no-langoptions are ignored and if neither is provided, the prompt asking the user wether they want to generate language statistics is skipped.Documentation
I updated the combined.svg used in the
READMEand removed the other two unused SVGs Also i changed theREADMEto reflect the new command line arguments. I also made some somewhat random changes here and there, the bigger one being in the usage instructions. These changes have various reasons behind them but are mostly subjective and stylistic, so have a look at them and let me know if the changes are appreciated.Output
I made some changes to the output the program produces:
git cloneoutput for cleaner interaction with the user