Broken links isssue tracker

[hchandad]

Overview

The old documentation was created using hugo, after the migration to contentlayer multiple links no longer point to the right resource, I tried to collect as many and list them, some have been previously mentioned in the issue tracker.

related: #360 #372 #370 #384 #363 #341 #189 #188

The table below was generated by running the linkchecker tool

$ linkchecker -o csv http://localhost:3000 1>link-checker.csv

The csv output was further converted to a markdown format

URL List

URL	Parent URL	Real URL	Fixed
/fonts/Inter.woff2	Parent URL	Real URL
/community/team	Parent URL	Real URL
/blog/2024-02-16-unikraft-releases-v0.16.2	Parent URL	Real URL
/docs/usage/advanced/kconfig/	Parent URL	Real URL	#389
/docs/develop/booting/	Parent URL	Real URL	#389
/docs/cli/reference/kraft/pkg/ls	Parent URL	Real URL	#389
/docs/cli/reference/kraft/pkg/rm	Parent URL	Real URL	#389
/docs/cli/reference/kraft/rm	Parent URL	Real URL	#389
/docs/cli/reference/kraft/net/ls	Parent URL	Real URL	#389
/docs/cli/reference/kraft/net/rm	Parent URL	Real URL	#389
/docs/contributing/kraftkit	Parent URL	Real URL	#380
docs	Parent URL	Real URL	#389
/docs/internals/testing	Parent URL	Real URL	#389
/guides/catalog-internals	Parent URL	Real URL	#389
unikraft.org/docs/contributing/review-process/	Parent URL	Real URL
/assets/files/eurosys2021-slides.pdf	Parent URL	Real URL	#389
Multiple Image Assets	Parent URL	Real URL	#389
%5B#link-to-commit%5D(unikraft/unikraft@142e842)	Parent URL	Real URL
/assets/imgs/unikraft-arch.jpg	Parent URL	Real URL	#389
/docs/operations/plats/kvm/	Parent URL	Real URL
/docs/operations/plats/xen/	Parent URL	Real URL
/docs/operations/plats/linuxu/	Parent URL	Real URL
/docs/cli/rootfs	Parent URL	Real URL	#389
docs/develop/porting/#makefileuk	Parent URL	Real URL	#389
/docs/cli/reference/kraft/cloud/deploy	Parent URL	Real URL
/guides/bincompat	Parent URL	Real URL
docs/contributing/suggest-changes	Parent URL	Real URL	#389
docs/contributing/coding-conventions#definitions	Parent URL	Real URL	#389

Getting the context of where a link is used

To figure out where the url is referenced in the files, git grep is useful , for example :

git grep -n 'docs/contributing/coding-conventions#definitions'
content/docs/contributing/coding-conventions.mdx:748:This is referenced in [the definitions](docs/contributing/coding-conventions#definitions) section for prefixes (i.e. the use of `uk_`, `ukplat_`, `ukarch_` prefixes).

Converting the csv output to markdown table

The following python script was used

convert.py

if __name__ == "__main__":
    import argparse
    import csv

    parser = argparse.ArgumentParser()
    parser.add_argument("-f", "--file", type=argparse.FileType())

    args = parser.parse_args()

    Columns = (
        "urlname",
        "parentname",
        "base",
        "result",
        "warningstring",
        "infostring",
        "valid",
        "url",
        "line",
        "column",
        "name",
        "dltime",
        "size",
        "checktime",
        "cached",
        "level",
        "modified",
    )

    def skip(iterator, n):
        for i in range(n):
            next(iterator)

    import sys

    if args.file:
        csvreader = csv.reader(args.file, delimiter=';')
        skip(csvreader, 4)
        for row in csvreader:
            line = {Columns[i]: value for i, value in enumerate(row)}
            try:
                mk_row = f"|{line['urlname']}|[Parent URL]({line['parentname']})|[Real URL]({line['url']})| |"
                print(mk_row)
            except KeyError:
                print(line, file=sys.stderr)

Further notes

• It would be nice to integrate a URL checking tool into the development workflow e.i:
  • Add an entry in the contributing docs detailing how to run a url checker on the changes made as part of the pull requests process
  • Integrate a url checker tool, as a pre-commit hook , it would be nice if we can determine which pages are affected from a commit diff and run the checker on those page's only, this would allow for faster run times
• some url checker tools already provide github action definitions, for example lychee which support's markdown directly.

[hchandad]

In regards to:

• /docs/operations/plats/kvm/
• /docs/operations/plats/xen/
• /docs/operations/plats/linuxu/

These exist original in the old tree, they are mostly written with the old version of kraft in mind. Thus if they are to be imported they should be rewritten.

There is an existing issue tracking the support of different hypervisors in kraftkit.

[hchandad]

For static media, they mostly exist withing the repo, under the static/assets/ folder, but the current static content is served from public/ instead, we can either cp the files to the public/ folder or configure the server to serve content from the old static folder as well.

Example:

• For this page images are under /static/assets/imgs/ but not loaded
• same for file assets /assets/files/eurosys2021-slides.pdf