#224 Fedora_docs_frontpage_redesign
Opened 2 years ago by ladyluchi. Modified 2 years ago
fedora-docs/ ladyluchi/docs-fp-o prod  into  master

Add package maintainers
Otto Urpelainen • 2 years ago  
Move eln-docs to GitHub
Aleksandra Fedorova • 3 years ago  
Publish PgM docs to prod
Ben Cotton • 4 years ago  
file modified
+1
@@ -1,3 +1,4 @@ 

  build

  public

  cache

+ build-scripts/output

file modified
+3 -3
@@ -1,9 +1,9 @@ 

- FROM fedora:29

+ FROM fedora:34

  

- RUN dnf -y module install nodejs:10 && \

+ RUN dnf -y module install nodejs:16 && \

      dnf clean all

  

- RUN npm i -g @antora/cli@2.0 @antora/site-generator-default@2.0

+ RUN npm i -g @antora/cli@3.0.0 @antora/site-generator@3.0.0 @antora/lunr-extension@1.0.0-alpha.6

  

  WORKDIR /antora

  ENTRYPOINT [ "antora" ]

file modified
+6 -6
@@ -1,13 +1,13 @@ 

- FROM fedora:29

+ FROM fedora:34

  

- RUN dnf -y module install nodejs:10/default && \

-     dnf -y install git && \

+ RUN dnf -y module install nodejs:16 && \

+     dnf -y install git python3-yaml && \

      dnf clean all

  

- RUN npm i -g @antora/cli@2.0 @antora/site-generator-default@2.0

+ RUN npm i -g @antora/cli@3.0.0 @antora/site-generator@3.0.0 @antora/lunr-extension@1.0.0-alpha.6

  

- ADD rebuild-site.sh /antora/rebuild-site.sh

+ ADD rebuild-site.py /antora/rebuild-site.py

  

  WORKDIR /antora

  

- CMD /antora/rebuild-site.sh

+ CMD /antora/rebuild-site.py

file modified
+41 -2
@@ -1,7 +1,46 @@ 

  # Building the docs

  

+ This script is used in OpenShift to rebuild the whole site for Fedora.

+ 

+ 

+ ## Local testing

+ 

+ Test the build locally using podman.

+ 

+ First, make sure your in the right directory — it's the same one that has this *README.md* file. Then build the container on your system:

+ 

+ ```

+ $ podman build --no-cache -t fedora-docs-openshift-rebuild . 

+ ```

+ 

+ And then run it:

+ 

+ ```

+ $ mkdir output

+ $ podman run --rm -it -e "BUILD_ENV=prod" -v $(pwd)/output:/antora/output:Z fedora-docs-openshift-rebuild

+ ```

+ 

+ The output will be saved in `./output/`.

+ 

+ ### Changing configuration

+ 

+ You can also set different environment variables by setting `-e "VAR_NAME=value"` in the `podman run` command.

+ See a list of supported variables below.

+ 

  ```

- docker run --rm -it -e "BUILD_ENV=stg" -v $(pwd)/output:/antora/output antora-openshift-rebuild

+ # The script requires the following environment variables to be set:

+ # 

+ # Variable          Values      Effect

+ # -----------------|-----------------------------------------------------------

+ # BUILD_ENV         prod        Build the production version of the site

+ #                               using the "prod" branch. DEFAULT

+ #                   stg         Build the staging version of the site

+ #                               using the "stg" branch.

+ # -----------------|-----------------------------------------------------------

+ # BUILD_LANGS       english     Build the "en-US" version only. DEFAULT

+ #                   translated  Build only the translated versions.

+ #                   all         Build the "en-US" and the translated versions.

  ```

  

- The output will be in `./output/`.

+ Please note that rebuilding all the languages locally will take many hours and results in gigabytes of output.

+ 

@@ -0,0 +1,575 @@ 

+ #!/usr/bin/python3

+ 

+ 

+ # The script requires the following environment variables to be set:

+ # 

+ # Variable          Values      Effect

+ # -----------------|-----------------------------------------------------------

+ # BUILD_ENV         prod        Build the production version of the site

+ #                               using the "prod" branch. DEFAULT

+ #                   stg         Build the staging version of the site

+ #                               using the "stg" branch.

+ # -----------------|-----------------------------------------------------------

+ # BUILD_LANGS       english     Build the "en-US" version only. DEFAULT

+ #                   translated  Build only the translated versions.

+ #                   all         Build the "en-US" and the translated versions.

+ #

+ #

+ 

+ 

+ import tempfile

+ import yaml

+ import os

+ import errno

+ import subprocess

+ import copy

+ import datetime

+ import shutil

+ import sys

+ import glob

+ import re

+ 

+ 

+ def get_config():

+     config = {}

+ 

+     config["docs_repo_branch"] = os.getenv(

+         "BUILD_ENV",

+         "prod")

+ 

+     config["build_langs"] = os.getenv(

+         "BUILD_LANGS",

+         "english")

+ 

+     config["docs_repo_url"] = os.getenv(

+         "DOCS_REPO_URL",

+         "https://pagure.io/fedora-docs/docs-fp-o.git")

+ 

+     config["translated_sources_repo_url"] = os.getenv(

+         "TRANSLATED_SOURCES_REPO_URL",

+         "https://pagure.io/fedora-docs/translated-sources.git"

+     )

+ 

+     config["translated_adockeywords_repo_url"] = os.getenv(

+         "TRANSLATED_ADOCKEYWORDS_REPO_URL",

+         "https://pagure.io/fedora-docs-l10n/asciidoc-keywords.git")

+     

+     os.environ['CI'] = "true"

+ 

+     return config

+ 

+ 

+ def log(msg):

+     print(msg, flush=True)

+ 

+ 

+ def get_languages(config):

+     languages_dict = {}

+ 

+     with tempfile.TemporaryDirectory() as workdir:

+         translated_sources_repo = os.path.join(workdir, "translated_sources_repo")

+         subprocess.run(["git", "clone", "--depth=1", config["translated_sources_repo_url"], translated_sources_repo])

+ 

+         languages = []

+         filename_blacklist = [".git"]

+         for filename in os.listdir(translated_sources_repo):

+             filepath = os.path.join(translated_sources_repo, filename)

+             if os.path.isdir(filepath) and filename not in filename_blacklist:

+                 languages.append(filename)

+ 

+         languages.sort()

+ 

+         for lang in languages:

+             languages_dict[lang] = []

+ 

+             lang_dir = os.path.join(translated_sources_repo, lang)

+             for component in os.listdir(lang_dir):

+                 version_dir = os.path.join(lang_dir, component)

+                 for version in os.listdir(version_dir):

+                     start_path = "{lang}/{component}/{version}".format(

+                             lang=lang, component=component, version=version)

+ 

+                     # This is a workaround for cases when a component doesn't have

+                     # the ROOT module that should contain the antora.yml.

+                     # Without the ROOT module the translation scripts won't

+                     # pick up the antora.yml causing the docs build to fail.

+                     # So we're only including directories with the antora.yml file.

+                     if "antora.yml" in os.listdir(os.path.join(version_dir, version)):

+                         languages_dict[lang].append(start_path)

+ 

+     return languages_dict

+ 

+ 

+ def generate_lang_switch_ui(languages):

+     template_start = """

+         <div class="page-languages">

+             <button class="languages-menu-toggle" title="Show other languages of the site">

+                 {{{env.ANTORA_LANGUAGE}}}

+             </button>

+             <div class="languages-menu">

+     """

+     template_end = """

+             </div>

+         </div>

+     """

+ 

+     template_list = []

+     template_list.append(template_start)

+     for language in languages:

+         link = '<a class="language" href="{{{{siteRootPath}}}}/../{language}{{{{page.url}}}}">{language}</a>'.format(

+                     language=language)

+         template_list.append(link)

+     template_list.append(template_end)

+ 

+     return "\n".join(template_list)

+ 

+ def get_metadata_from_branch(branch, repodir):

+     cmd = f"""

+       for file in $(git ls-tree -r --name-only origin/{branch}|grep -E "adoc$"); do 

+         git log -1 --format="'$file':%n  refs: %S%n  hash: %h%n  date: %as" origin/{branch} -- $file;

+       done"""

+     s = subprocess.run(cmd, shell=True, cwd=repodir,  stdout=subprocess.PIPE, stderr=subprocess.STDOUT)

+     return yaml.safe_load(s.stdout)

+ 

+ def prefetch_sources(original_sources, site_yml, config):

+     if not os.path.isdir(original_sources):

+         os.mkdir(original_sources)

+ 

+     default_branch = site_yml["content"].get("branches", ["master"])

+     if type(default_branch) is str:

+         default_branch = [default_branch]

+ 

+     metadata = {}

+ 

+     for repo_data in site_yml["content"]["sources"]:

+       gitdir = tempfile.TemporaryDirectory(dir=original_sources)

+       dirname = gitdir.name

+       gitdir.cleanup()

+       log(f"Fetching {repo_data['url']}...")

+       subprocess.run(["git", "clone", repo_data['url'], dirname])

+       repo_data['url'] = dirname

+ 

+       branches = repo_data.get("branches", default_branch)

+       if type(branches) is str:

+         branches = [branches]

+ 

+       for branch in branches:

+         y = get_metadata_from_branch(branch, repo_data['url'])

+         if dirname in metadata:

+           metadata[dirname][branch] = y

+         else:

+           metadata[dirname] = { branch: y }

+ 

+     if metadata:

+       with open("/tmp/metadata.yml", "w") as f:

+         yaml.dump(metadata, f)

+ 

+     return site_yml

+ 

+ def prepare_translated_sources(translated_sources, site_yml, languages, config):

+     with tempfile.TemporaryDirectory() as workdir:

+         # Location of the translated-sources repo

+         #   (This repo only holds content that has some translations done,

+         #   so it might be incomplete.)

+         translated_sources_original = os.path.join(workdir, "translated_sources_original")

+ 

+         # Location of the English sources in the same structure

+         # as translated-sources, so:

+         #    COMPONENT/VERSION/antora.yml

+         #    COMPONENT/VERSION/modules/MODULE1

+         #    COMPONENT/VERSION/modules/MODULE2

+         #    COMPONENT/VERSION/modules/...

+         en_sources = os.path.join(workdir, "en_sources")

+ 

+         # Clone the original translated-sources

+         subprocess.run(["git", "clone", "--depth=1", config["translated_sources_repo_url"], translated_sources_original])

+ 

+         # Get a list of the original English repos

+         repos = []

+         default_branch = site_yml["content"].get("branches", ["master"])

+         if type(default_branch) is str:

+             default_branch = [default_branch]

+ 

+         for repo_data in site_yml["content"]["sources"]:

+ 

+             branches = repo_data.get("branches", default_branch)

+             if type(branches) is str:

+                 branches = [branches]

+ 

+             start_path = repo_data.get("start_path", "")

+ 

+             for branch in branches:

+                 repo = {}

+                 repo["url"] = repo_data["url"]

+                 repo["branch"] = branch

+                 repo["start_path"] = start_path

+ 

+                 repos.append(repo)

+         

+         # Clome the original English sources and put them into the

+         # desired structure in en_sources (described above)

+         components = {}

+         for repo in repos:

+             with tempfile.TemporaryDirectory() as tmp_repo_root:

+                 log("")

+                 log("Cloning {url} {branch}".format(url=repo["url"], branch=repo["branch"]))

+                 subprocess.run(["git", "clone", "--branch", repo["branch"], "--depth=1", repo["url"], tmp_repo_root])

+ 

+                 repo_dir = os.path.join(tmp_repo_root, repo["start_path"])

+                 antora_yml_file = os.path.join(repo_dir, "antora.yml")

+                 with open(antora_yml_file, "r") as file:

+                     antora_yml = yaml.safe_load(file)

+ 

+                 component = antora_yml["name"]

+                 version = antora_yml["version"]

+                 # if component version is null (~), fallback to "master"

+                 if not version:

+                     version = "master"

+ 

+                 # Saving components and all their versions

+                 if component not in components:

+                     components[component] = set()

+                 components[component].add(version)

+ 

+                 for module in os.listdir(os.path.join(repo_dir, "modules")):

+                     # Now this is looping over modules accross all components,

+                     # so component, module, and version variables are available

+ 

+                     try:

+                         os.makedirs(os.path.join(en_sources, component, version, "modules"))

+                     except OSError as e:

+                         if e.errno != errno.EEXIST:

+                             raise

+                     

+                     original_module_dir = os.path.join(repo_dir, "modules", module)

+                     module_dir = os.path.join(en_sources, component, version, "modules", module)

+ 

+                     subprocess.run(["cp", "-a", original_module_dir, module_dir])

+ 

+                     if module == "ROOT" or "nav" in antora_yml:

+                         # if this is the main antora.yml file

+                         subprocess.run(["cp", "-a", antora_yml_file, os.path.join(en_sources, component, version, "antora.yml")])

+                         log("----- copying antora.yml for  {component} {module} {version}".format(

+                                 component=component, module=module, version=version))

+                     else:

+                         log("----- skipping antora.yml for  {component} {module} {version}".format(

+                                 component=component, module=module, version=version))

+ 

+         # Set up the language structure in translated_sources

+         for language in languages:

+             lang_dir = os.path.join(translated_sources, language)

+             try:

+                 os.makedirs(lang_dir)

+             except OSError as e:

+                 if e.errno != errno.EEXIST:

+                     raise

+         

+         # Copy the English sources in the translated_sources

+         for language in languages:

+             lang_dir = os.path.join(translated_sources, language)

+             for component in components:

+                 en_component_dir = os.path.join(en_sources, component)

+                 subprocess.run(["cp", "-a", en_component_dir, lang_dir])

+ 

+         # And finally copy the original translated sources

+         # into translated_sources

+         for language in languages:

+             src = os.path.join(translated_sources_original, language)

+             subprocess.run(["cp", "-a", src, translated_sources + "/"])

+ 

+     return components

+ 

+ 

+ def prepare_localized_admonitions(languages, config):

+     """ Asciidoc use keywords for admonitions and others items """

+     keywords = {}

+ 

+     with tempfile.TemporaryDirectory() as workdir:

+         translated_keywords_repo = os.path.join(workdir, "asciidoc-keywords")

+         subprocess.run(["git", "clone", "--depth=1", config["translated_adockeywords_repo_url"], translated_keywords_repo])

+ 

+         languages = []

+         log(translated_keywords_repo + "/langs/*/asciidoc-attributes.yml")

+         for filename in glob.glob(translated_keywords_repo + "/langs/*/asciidoc-attributes.yml"):

+             languages.append(filename.rsplit("/")[::-1][1])

+ 

+         for lang in languages:

+             file = translated_keywords_repo + "/langs/" + lang + "/asciidoc-attributes.yml"

+             with open(file, 'r') as stream:

+                 try:

+                     keywords[lang] = yaml.load(stream)

+                 except yaml.YAMLError as exc:

+                     print(exc)

+ 

+     return keywords

+ 

+ 

+ def init_git_repo(path):

+     try:

+         os.makedirs(path)

+     except OSError as e:

+         if e.errno != errno.EEXIST:

+             raise

+     subprocess.run(["git", "init"], cwd=path)

+     subprocess.run(["git", "config", "user.name", "Your Name"], cwd=path)

+     subprocess.run(["git", "config", "user.email", "you@example.com"], cwd=path)

+     subprocess.run(["git", "commit", "--allow-empty", "-m", "init"], cwd=path)

+ 

+ def get_version_from_file(index_file):

+     latest = None

+     if os.path.isfile(index_file):

+       with open(index_file, 'r') as f:

+         for l in f:

+           m = re.search(r'data-version="(f\d+)"', l)

+           if m:

+             latest = m[1]

+             break

+     return latest

+ 

+ 

+ def gen_redirect(lang, results_dir):

+     htaccess = os.path.join(results_dir, ".htaccess")

+     latest = get_version_from_file(f"{results_dir}/fedora/latest/index.html")

+     if latest:

+       htaccess_content = f"""

+         ErrorDocument 404 /{lang}/404.html

+         Redirect 302 /{lang}/fedora/{latest} /{lang}/fedora/latest

+         """

+       with open(htaccess, "w") as file:

+           file.write(htaccess_content)

+ 

+ def main():

+     config = get_config()

+     

+     with tempfile.TemporaryDirectory() as workdir:

+ 

+         #####--------------------------------------#####

+         #####  Preparation                         #####

+         #####--------------------------------------#####

+ 

+         # Location of the docs-fp-o repo

+         docs_repo = os.path.join(workdir, "docs_repo")

+ 

+         # Location of the translated sources used for the build

+         #   (That's the translated-sources repo with the missing files added

+         #   from the English sources.)

+         translated_sources = os.path.join(workdir, "translated_sources")

+         init_git_repo(translated_sources)

+ 

+         log("")

+         log("===== Getting the site definition (site.yml) =====")

+         subprocess.run(["git", "clone", "--branch", config["docs_repo_branch"], "--depth=1", config["docs_repo_url"], docs_repo])

+ 

+         with open(os.path.join(docs_repo, "site.yml"), "r") as file:

+             original_site_yml = yaml.safe_load(file)

+ 

+         log("")

+         log("===== Getting a list of languages =====")

+         languages = get_languages(config)

+         log("  Languages: {}".format(" ".join(languages)))

+ 

+         log("")

+         log("===== Generating the language switch UI =====")

+         lang_switch_ui = generate_lang_switch_ui(["en-US"] + list(languages.keys()))

+ 

+         ui_dir = os.path.join(docs_repo, "supplemental-ui", "partials")

+         try:

+             os.makedirs(ui_dir)

+         except OSError as e:

+             if e.errno != errno.EEXIST:

+                 raise

+ 

+         ui_file = os.path.join(ui_dir, "page-languages.hbs")

+         with open(ui_file, "w") as file:

+             file.write(lang_switch_ui)

+ 

+         # Timestamp to be included in the footer of the docs

+         timestamp = datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S UTC')

+         antora_env = copy.deepcopy(os.environ)

+         antora_env["ANTORA_DATE"] = timestamp

+ 

+ 

+ 

+         #####--------------------------------------#####

+         #####  English site build                  #####

+         #####--------------------------------------#####

+ 

+         if config["build_langs"] == "english" or config["build_langs"] == "all":

+ 

+             log("")

+             log("===== Prefetching all components =====")

+             original_sources = os.path.join(workdir, "orginal_sources")

+             site_local = prefetch_sources(original_sources, original_site_yml, config)

+             log(original_site_yml["content"]["sources"][0]['url'])

+             log(site_local["content"]["sources"][0]['url'])

+ 

+             filename = os.path.join(docs_repo, "site-local.yml")

+             with open(filename, "w") as file:

+                 file.write(yaml.dump(site_local))

+ 

+             

+ 

+             log("")

+             log("===== Building the en-US site =====")

+             antora_env["ANTORA_LANGUAGE"] = "en-US"

+             # result = subprocess.run(["antora", "--html-url-extension-style=indexify", os.path.join(docs_repo, "site.yml")], env=antora_env)

+             result = subprocess.run(["antora", "--extension=lunr", filename], env=antora_env)

+ 

+             if result.returncode != 0:

+                 log("ERROR building the en-US site")

+                 sys.exit(1)

+                 

+             log("")

+             log("===== Copying the en-US site =====")

+             source_dir = os.path.join(docs_repo, "public")

+             target_dir_copying = "/antora/output/en-US.building"

+             target_dir_copied = "/antora/output/en-US.building/en-US"

+             target_dir_final = "/antora/output/en-US"

+ 

+             # Generate htaccess for redirect

+             gen_redirect("en-US", os.path.join(source_dir, "en-US"))

+ 

+             shutil.rmtree(target_dir_copying, ignore_errors=True)

+             subprocess.run(["cp", "-a", source_dir, target_dir_copying])

+             shutil.rmtree(target_dir_final, ignore_errors=True)

+             shutil.move(target_dir_copied, target_dir_final)

+             shutil.rmtree(target_dir_copying, ignore_errors=True)

+ 

+             log("")

+             log("===== Copying the index.html =====")

+             source_index_html = os.path.join(docs_repo, "static", "index.html")

+             target_index_html = "/antora/output/index.html"

+             shutil.copy(source_index_html, target_index_html)

+ 

+ 

+ 

+         #####--------------------------------------#####

+         #####  Translated site build               #####

+         #####--------------------------------------#####

+ 

+ 

+         if config["build_langs"] == "translated" or config["build_langs"] == "all":

+ 

+             log("")

+             log("===== Preparing the translated sources =====")

+             components = prepare_translated_sources(translated_sources, original_site_yml, languages, config)

+ 

+             log("")

+             log("===== Preparing the translated ascidoc keywords =====")

+             keywords = prepare_localized_admonitions(languages, config)

+ 

+             log("")

+             log("===== Generating site.lang.yml files =====")

+             for lang in languages:

+                 lang_site_yml = copy.deepcopy(original_site_yml)

+ 

+                 lang_site_yml["output"]["dir"] = "./public/{lang}".format(lang=lang)

+ 

+                 lang_site_yml["content"] = {}

+                 # Branches are set to HEAD because the script uses a locally-generated

+                 # content on top of the original translated-sources repo

+                 lang_site_yml["content"]["branches"] = "HEAD"

+                 # disable the "Edit this Page"

+                 lang_site_yml["content"]["edit_url"] = False

+                 lang_site_yml["content"]["sources"] = []

+ 

+                 for component, versions in components.items():

+                     for version in versions:

+                         source = {}

+                         source["url"] = translated_sources

+                         source["start_path"] = "{lang}/{component}/{version}".format(

+                                     lang=lang, component=component, version=version)

+ 

+                         lang_site_yml["content"]["sources"].append(source)

+ 

+                 if lang in keywords:

+                     if "attributes" in lang_site_yml["asciidoc"]:

+                         lang_site_yml["asciidoc"]["attributes"].append(keywords[lang])

+                     else:

+                         lang_site_yml["asciidoc"]["attributes"] = keywords[lang]

+ 

+                 # disable page-last-update extension 

+                 if "antora" in lang_site_yml and "extensions" in lang_site_yml["antora"]:

+                   for ext in lang_site_yml["antora"]["extensions"]:

+                     if 'id' in ext and ext['id'] == "page-last-update":

+                       ext['enabled'] = False

+ 

+                 # don't pull ui-bundle (use cache)

+                 lang_site_yml["runtime"]["fetch"] = False

+ 

+                 filename = os.path.join(docs_repo, "site-{lang}.yml".format(lang=lang))

+                 with open(filename, "w") as file:

+                     file.write(yaml.dump(lang_site_yml))

+ 

+                 log("  {lang} done".format(lang=lang))

+ 

+             # Building all the translated sites

+             lastlang = list(languages)[-1]

+             for lang in languages:

+                 log("")

+                 log("===== Building the {lang} site =====".format(lang=lang))

+                 filename = "site-{lang}.yml".format(lang=lang)

+                 antora_env["ANTORA_LANGUAGE"] = lang

+                 result = subprocess.run(["antora", "--extension=lunr", os.path.join(docs_repo, filename)], env=antora_env)

+ 

+                 if result.returncode != 0:

+                     log("ERROR building the {lang} site".format(lang=lang))

+                     sys.exit(1)

+ 

+                 # Copying the translated site

+                 log("")

+                 log(f"=== Copying the {lang} site ===")

+ 

+                 results_dir = os.path.join(docs_repo, "public", lang)

+                 publish_dir = f"/antora/output/{lang}"

+                 copying_dir = f"/antora/output/{lang}.tmp"

+                 lastlang_dir = f"/antora/output/{lastlang}"

+ 

+                 # I have:

+                 # docs_repo/public/xx         <- results_dir  (local partition)

+                 # /antora/output/xx.tmp       <- copying_dir  (mounted partition)

+                 # /antora/output/xx           <- publish_dir  (mounted partition)

+                 #

+                 # I need to:

+                 #   1/  copy from local to mounted

+                 #   2/  remove old in mounted

+                 #   3/  move new within mounted

+                 #   4/  remove the copying dir

+ 

+                 # Generate htaccess for redirect

+                 gen_redirect(lang, results_dir)

+ 

+                 # Make sure copying_dir doesn't exist

+                 shutil.rmtree(copying_dir, ignore_errors=True)

+ 

+                 # Copy results from local partition to a mounted partition

+                 log(f"Copying from {results_dir} to {copying_dir}")

+                 subprocess.run(["cp", "-a", results_dir, copying_dir])

+ 

+                 # Swap the old tree for the new one for each language

+                 log(f"Moving language to the final place: {copying_dir} to {publish_dir}")

+                 shutil.rmtree(publish_dir, ignore_errors=True)

+                 subprocess.run(["mv", copying_dir, publish_dir])

+ 

+                 # Recreate hardlinks as we go, in case the rsync job

+                 # start while we are still building

+                 if (lang != lastlang):

+                     log(f"hardlinking files: {publish_dir} - {lastlang_dir}")

+                     subprocess.run(["hardlink", "-cv", publish_dir, lastlang_dir])

+ 

+                 # Remove local build

+                 shutil.rmtree(results_dir, ignore_errors=True)

+ 

+             # End Building all the translated sites

+ 

+             # https://pagure.io/fedora-infrastructure/issue/8964

+             log("Consolidate files with hardlink...")

+             subprocess.run(["hardlink", "-cv", "/antora/output/"])

+ 

+             log("DONE!")

+     

+ 

+ 

+ if __name__ == "__main__":

+     main()

+ 

@@ -1,59 +0,0 @@ 

- #/bin/bash

- 

- echo ""

- echo "$(date +%G-%m-%d-%H%M%S): Building Fedora Docs..."

- echo ""

- 

- SOURCE_REPO="https://pagure.io/fedora-docs/docs-fp-o.git"

- BUILD_CMD="antora --html-url-extension-style=indexify site.yml"

- 

- target="/antora"

- 

- if [ -z ${BUILD_ENV+x} ]; then 

-     BUILD_ENV="master"   

- fi

- 

- 

- if ! workdir=$(mktemp -d) ; then

-     echo "Error creating a tempdir. Exiting."

-     exit 1

- fi

- 

- if ! git clone --single-branch --branch $BUILD_ENV  $SOURCE_REPO "$workdir/source" ; then

-     echo ""

-     echo "Error cloning the source repo."

-     rm -rf "$workdir"

-     exit 1

- else

-     echo "OK  cloned source repo"

- fi

- 

- if [ ! -f "$workdir/source/site.yml" ]; then

-     echo ""

-     echo "Error! There is no site.yml in the source."

-     rm -rf "$workdir"

-     exit 1

- fi

- 

- pushd "$workdir/source/" > /dev/null

- 

- 

- export ANTORA_DATE=$(date -u)

- 

- if ! $BUILD_CMD ; then

-     echo ""

-     echo "Error building the site."

-     rm -rf "$workdir"

-     exit 1

- fi

- 

- popd > /dev/null

- 

- cp -r "$workdir/source/public/en-US" "$target/output/en-US.building" && rm -rf "$target/output/en-US" && mv "$target/output/en-US.building" "$target/output/en-US" || exit 1

- cp "$workdir/source/static/index.html" "$target/output/index.html" || exit 1

- 

- rm -rf "$workdir"

- 

- echo ""

- echo "$(date +%G-%m-%d-%H%M%S): Building of Fedora Docs has finished successfully!"

- echo ""

file modified
+2 -2
@@ -1,7 +1,7 @@ 

  #!/bin/sh

  

- image="docker.io/antora/antora"

- cmd="--html-url-extension-style=indexify site.yml"

+ image="docker.io/antora/antora:3.0.0"

+ cmd="site.yml"

  

  if [ "$(uname)" == "Darwin" ]; then

      # Running on macOS.

@@ -0,0 +1,38 @@ 

+ 

+ module.exports.register = function () {

+   this.on('contentClassified', async ({ contentCatalog }) => {

+     const logger = this.require('@antora/logger').get('page-last-update')

+     const yaml = this.require('js-yaml')

+     const path = this.require('path')

+     const fs = this.require('fs')

+     const pages = contentCatalog.findBy({ family: 'page' })

+     const pageRx = new RegExp('modules/[^/]+/pages/(?:.+/)*[^.].*\.adoc$')

+ 

+     if (fs.existsSync("/tmp/metadata.yml")) {

+       let data = fs.readFileSync("/tmp/metadata.yml", "utf-8")

+       let y = yaml.loadAll(data)

+     

+       for (const page of pages) {

+         if (page.src.origin.type == "git" && pageRx.test(page.src.path)) {

+           const relpath = path.join(page.src.origin.startPath, page.src.path)

+           const abspath = page.src.origin.gitdir.replace(/\/\.git$/, "")

+           // logger.warn(abspath)

+           const contents = page.contents.toString('utf-8')

+           const branch = page.src.origin.branch

+           if (abspath in y[0] && branch in y[0][abspath] && relpath in y[0][abspath][branch]) {

+             let last_update = y[0][abspath][branch][relpath]["date"].toISOString().split('T')[0]

+             logger.warn({name: relpath, rev: branch, update: last_update})

+             

+             // Inject attribute in page content

+             const patchedContents = ":page-last-updated: " + last_update + "\n" + contents

+             page.contents = Buffer.from(patchedContents)

+           }

+         }

+       }

+ 

+     } else {

+       logger.warn("no git metadata found, skipping...")

+     }

+     this.updateVariables({ contentCatalog })

+   })

+ }

file modified
+99 -10
@@ -1,5 +1,5 @@ 

  site:

-   title: Fedora Docs Site

+   title: Fedora Docs

    url: https://docs.fedoraproject.org/en-US/

    start_page: docs::index.adoc

  content:
@@ -14,6 +14,11 @@ 

    - url: https://pagure.io/fedora-docs/release-docs-home.git

      branches:

      - master

+     - f35

+     - f34

+     - f33

+     - f32

+     - f31

      - f30

      - f29

      - f28
@@ -22,6 +27,11 @@ 

    - url: https://pagure.io/fedora-docs/install-guide.git

      branches:

      - master

+     - f35

+     - f34

+     - f33

+     - f32

+     - f31

      - f30

      - f29

      - f28
@@ -30,6 +40,11 @@ 

    - url: https://pagure.io/fedora-docs/system-administrators-guide.git

      branches:

      - master

+     - f35

+     - f34

+     - f33

+     - f32

+     - f31

      - f30

      - f29

      - f28
@@ -38,6 +53,11 @@ 

    - url: https://pagure.io/fedora-docs/release-notes.git

      branches:

      - master

+     - f35

+     - f34

+     - f33

+     - f32

+     - f31

      - f30

      - f29

      - f28
@@ -47,13 +67,26 @@ 

    - url: https://pagure.io/mentored-projects.git

    - url: https://pagure.io/fedora-commops.git

      start_path: docs

+     branches: main

+   - url: https://pagure.io/fedora-websites.git

+     start_path: docs

+     branches: main

    - url: https://pagure.io/Fedora-Council/council-docs.git

      start_path: council

+     branches: main

    - url: https://pagure.io/Fedora-Council/council-docs.git

      start_path: project

+     branches: main

+   - url: https://pagure.io/Fedora-Council/status_reports.git

+     branches: main

    - url: https://pagure.io/fedora-docs/modularity.git

+   - url: https://pagure.io/fedora-docs/package-maintainer-docs.git

+     branches: main

    - url: https://github.com/fedora-silverblue/silverblue-docs.git

-   - url: https://pagure.io/fedora-docs/documentation-contributors-guide.git

+   - url: https://pagure.io/fedora-kde/kinoite-docs

+     branches: main

+   - url: https://gitlab.com/fedora/docs/community-tools/documentation-contributors-guide.git

+     branches: main

    - url: https://pagure.io/fedora-docs/flatpak.git

    - url: https://pagure.io/fedora-diversity.git

    - url: https://pagure.io/packaging-committee.git
@@ -62,30 +95,86 @@ 

      start_path: website

    - url: https://pagure.io/fesco/fesco-docs.git

      start_path: fesco

-   - url: https://pagure.io/fedora-iot/iot-docs.git

-   - url: https://pagure.io/fedora-badges.git

-     start_path: docs

+   - url: https://github.com/fedora-iot/iot-docs.git

+     start_path: user-docs

+     branches: main

+   - url: https://github.com/fedora-iot/iot-docs.git

+     start_path: wg-docs

+     branches: main

+   - url: https://pagure.io/fedora-badges/docs.git

    - url: https://pagure.io/fedora-docs/remix-building.git

    - url: https://github.com/containers/docs.git

+     branches: main

    - url: https://pagure.io/neuro-sig/documentation.git

+     branches: main

    - url: https://pagure.io/sig-teleirc/infrastructure.git

      start_path: docs

    - url: https://pagure.io/fedora-ci/docs.git

    - url: https://pagure.io/java-packaging-howto.git

-   - url: https://pagure.io/fedora-docs/taiga-docs.git

+   - url: https://pagure.io/fedora-magazine.git

+   - url: https://pagure.io/minimization.git

+   - url: https://github.com/coreos/fedora-coreos-docs.git

+     branches: main

+   - url: https://pagure.io/cpe/rawhide-gating-docs.git

+   - url: https://pagure.io/cpe/docs.git

+   - url: https://pagure.io/fedora-qa/qa-docs.git

+   - url: https://pagure.io/fedora-pgm/pgm_docs.git

+     branches: main

+   - url: https://pagure.io/fedora-pgm/pgm_docs.git

+     branches: main

+     start_path: releases

+   - url: https://pagure.io/Ask-Fedora-SOP-docs.git

+   - url: https://pagure.io/fedora-join/fedora-join-docs.git

+   - url: https://pagure.io/fedora-docs/localization.git

+   - url: https://pagure.io/fedora-l10n/docs.git

+   - url: https://github.com/fedora-eln/eln-docs.git

+   - url: https://pagure.io/i3-sig/docs.git

+   - url: https://github.com/fedora-infra/fedora-accounts-docs.git

+     branches: main

+   - url: https://pagure.io/gaming/documentation.git

+     branches: main

+   - url: https://pagure.io/fedora-server.git

+     branches: main

+     start_path: docs

+   - url: https://pagure.io/fedora-server.git

+     branches: main

+     start_path: wg

+   - url: https://pagure.io/epel.git

+     branches: main

+   - url: https://pagure.io/infra-docs-fpo.git

+   - url: https://pagure.io/fedora-workstation/workstation-docs.git

+     branches: main

+   - url: https://gitlab.com/fedora/marketing/marketing-docs.git

+     branches: main

+   - url: https://pagure.io/defensive-coding-guide.git

  ui:

    bundle:

-     url: https://asamalik.fedorapeople.org/ui-bundle.zip

+     url: https://gitlab.com/fedora/docs/docs-website/ui-bundle/-/jobs/artifacts/HEAD/raw/build/ui-bundle.zip?job=bundle-stable

      snapshot: true

    default_layout: with_menu

+   supplemental_files: ./supplemental-ui

  output:

    clean: true

    dir: ./public/en-US

-   destinations:

-   - provider: archive

  runtime:

-   pull: true

+   fetch: true

    cache_dir: ./cache

+   log:

+     format: pretty

+     level: fatal

  asciidoc:

    extensions:

    - ./lib/extensions/package-inline-macro.js

+ antora:

+   extensions:

+   - id: page-last-update

+     require: ./lib/extensions/page-last-update.js

+     enabled: true

+   - id: lunr

+     require: '@antora/lunr-extension'

+     index_latest_only: true

+     enabled: false

+ urls:

+   latest_version_segment: latest

+   html_extension_style: indexify

+ 

@@ -0,0 +1,75 @@ 

+ .search-result-dropdown-menu {

+   position: absolute;

+   z-index: 100;

+   display: block;

+   right: 0;

+   left: inherit;

+   top: 100%;

+   border-radius: 4px;

+   margin: 6px 0 0;

+   padding: 0;

+   text-align: left;

+   height: auto;

+   background: transparent;

+   border: none;

+   max-width: 600px;

+   min-width: 500px;

+   box-shadow: 0 1px 0 0 rgba(0, 0, 0, 0.2), 0 2px 3px 0 rgba(0, 0, 0, 0.1);

+ }

+ 

+ @media screen and (max-width: 768px) {

+   .search-result-dropdown-menu {

+     min-width: calc(100vw - 3.75rem);

+   }

+ }

+ 

+ .search-result-dataset {

+   position: relative;

+   border: 1px solid #d9d9d9;

+   background: #fff;

+   border-radius: 4px;

+   overflow: auto;

+   padding: 0 8px;

+   max-height: calc(100vh - 5.25rem);

+   line-height: 1.5;

+ }

+ 

+ .search-result-item {

+   display: flex;

+   margin: 0.5rem 0;

+ }

+ 

+ .search-result-document-title {

+   width: 33%;

+   border-right: 1px solid #ddd;

+   color: #02060c;

+   font-weight: 500;

+   font-size: 0.8rem;

+   padding: 0.5rem 0.5rem 0.5rem 0;

+   text-align: right;

+   position: relative;

+   word-wrap: break-word;

+ }

+ 

+ .search-result-document-hit {

+   flex: 1;

+   font-size: 0.75rem;

+   color: #63676d;

+ }

+ 

+ .search-result-document-hit > a {

+   color: inherit;

+   display: block;

+   padding: 0.55rem 0.25rem 0.55rem 0.75rem;

+ }

+ 

+ .search-result-document-hit > a:hover {

+   background-color: rgba(69, 142, 225, 0.05);

+ }

+ 

+ .search-result-highlight {

+   color: #174d8c;

+   background: rgba(143, 187, 237, 0.1);

+   padding: 0.1em 0.05em;

+   font-weight: 500;

+ }

@@ -0,0 +1,248 @@ 

+ /* global CustomEvent */

+ ;(function (globalScope) {

+   /* eslint-disable no-var */

+   var config = document.getElementById('search-ui-script').dataset

+   var snippetLength = parseInt(config.snippetLength || 100, 10)

+   var siteRootPath = config.siteRootPath || ''

+   appendStylesheet(config.stylesheet)

+   var searchInput = document.getElementById('search-input')

+   var searchResult = document.createElement('div')

+   searchResult.classList.add('search-result-dropdown-menu')

+   searchInput.parentNode.appendChild(searchResult)

+ 

+   function appendStylesheet (href) {

+     if (!href) return

+     document.head.appendChild(Object.assign(document.createElement('link'), { rel: 'stylesheet', href: href }))

+   }

+ 

+   function highlightText (doc, position) {

+     var hits = []

+     var start = position[0]

+     var length = position[1]

+ 

+     var text = doc.text

+     var highlightSpan = document.createElement('span')

+     highlightSpan.classList.add('search-result-highlight')

+     highlightSpan.innerText = text.substr(start, length)

+ 

+     var end = start + length

+     var textEnd = text.length - 1

+     var contextAfter = end + snippetLength > textEnd ? textEnd : end + snippetLength

+     var contextBefore = start - snippetLength < 0 ? 0 : start - snippetLength

+     if (start === 0 && end === textEnd) {

+       hits.push(highlightSpan)

+     } else if (start === 0) {

+       hits.push(highlightSpan)

+       hits.push(document.createTextNode(text.substr(end, contextAfter)))

+     } else if (end === textEnd) {

+       hits.push(document.createTextNode(text.substr(0, start)))

+       hits.push(highlightSpan)

+     } else {

+       hits.push(document.createTextNode('...' + text.substr(contextBefore, start - contextBefore)))

+       hits.push(highlightSpan)

+       hits.push(document.createTextNode(text.substr(end, contextAfter - end) + '...'))

+     }

+     return hits

+   }

+ 

+   function highlightTitle (sectionTitle, doc, position) {

+     var hits = []

+     var start = position[0]

+     var length = position[1]

+ 

+     var highlightSpan = document.createElement('span')

+     highlightSpan.classList.add('search-result-highlight')

+     var title

+     if (sectionTitle) {

+       title = sectionTitle.text

+     } else {

+       title = doc.title

+     }

+     highlightSpan.innerText = title.substr(start, length)

+ 

+     var end = start + length

+     var titleEnd = title.length - 1

+     if (start === 0 && end === titleEnd) {

+       hits.push(highlightSpan)

+     } else if (start === 0) {

+       hits.push(highlightSpan)

+       hits.push(document.createTextNode(title.substr(length, titleEnd)))

+     } else if (end === titleEnd) {

+       hits.push(document.createTextNode(title.substr(0, start)))

+       hits.push(highlightSpan)

+     } else {

+       hits.push(document.createTextNode(title.substr(0, start)))

+       hits.push(highlightSpan)

+       hits.push(document.createTextNode(title.substr(end, titleEnd)))

+     }

+     return hits

+   }

+ 

+   function highlightHit (metadata, sectionTitle, doc) {

+     var hits = []

+     for (var token in metadata) {

+       var fields = metadata[token]

+       for (var field in fields) {

+         var positions = fields[field]

+         if (positions.position) {

+           var position = positions.position[0] // only higlight the first match

+           if (field === 'title') {

+             hits = highlightTitle(sectionTitle, doc, position)

+           } else if (field === 'text') {

+             hits = highlightText(doc, position)

+           }

+         }

+       }

+     }

+     return hits

+   }

+ 

+   function createSearchResult (result, store, searchResultDataset) {

+     result.forEach(function (item) {

+       var ids = item.ref.split('-')

+       var docId = ids[0]

+       var doc = store[docId]

+       var sectionTitle

+       if (ids.length > 1) {

+         var titleId = ids[1]

+         sectionTitle = doc.titles.filter(function (item) {

+           return String(item.id) === titleId

+         })[0]

+       }

+       var metadata = item.matchData.metadata

+       var hits = highlightHit(metadata, sectionTitle, doc)

+       searchResultDataset.appendChild(createSearchResultItem(doc, sectionTitle, item, hits))

+     })

+   }

+ 

+   function createSearchResultItem (doc, sectionTitle, item, hits) {

+     var documentTitle = document.createElement('div')

+     documentTitle.classList.add('search-result-document-title')

+     documentTitle.innerText = doc.title

+     var documentHit = document.createElement('div')

+     documentHit.classList.add('search-result-document-hit')

+     var documentHitLink = document.createElement('a')

+     documentHitLink.href = siteRootPath + doc.url + (sectionTitle ? '#' + sectionTitle.hash : '')

+     documentHit.appendChild(documentHitLink)

+     hits.forEach(function (hit) {

+       documentHitLink.appendChild(hit)

+     })

+     var searchResultItem = document.createElement('div')

+     searchResultItem.classList.add('search-result-item')

+     searchResultItem.appendChild(documentTitle)

+     searchResultItem.appendChild(documentHit)

+     searchResultItem.addEventListener('mousedown', function (e) {

+       e.preventDefault()

+     })

+     return searchResultItem

+   }

+ 

+   function createNoResult (text) {

+     var searchResultItem = document.createElement('div')

+     searchResultItem.classList.add('search-result-item')

+     var documentHit = document.createElement('div')

+     documentHit.classList.add('search-result-document-hit')

+     var message = document.createElement('strong')

+     message.innerText = 'No results found for query "' + text + '"'

+     documentHit.appendChild(message)

+     searchResultItem.appendChild(documentHit)

+     return searchResultItem

+   }

+ 

+   function clearSearchResults (reset) {

+     if (reset === true) searchInput.value = ''

+     searchResult.innerHTML = ''

+   }

+ 

+   function search (index, text) {

+     // execute an exact match search

+     var result = index.search(text)

+     if (result.length > 0) {

+       return result

+     }

+     // no result, use a begins with search

+     result = index.search(text + '*')

+     if (result.length > 0) {

+       return result

+     }

+     // no result, use a contains search

+     result = index.search('*' + text + '*')

+     return result

+   }

+ 

+   function searchIndex (index, store, text) {

+     clearSearchResults(false)

+     if (text.trim() === '') {

+       return

+     }

+     var result = search(index, text)

+     var searchResultDataset = document.createElement('div')

+     searchResultDataset.classList.add('search-result-dataset')

+     searchResult.appendChild(searchResultDataset)

+     if (result.length > 0) {

+       createSearchResult(result, store, searchResultDataset)

+     } else {

+       searchResultDataset.appendChild(createNoResult(text))

+     }

+   }

+ 

+   function confineEvent (e) {

+     e.stopPropagation()

+   }

+ 

+   function debounce (func, wait, immediate) {

+     var timeout

+     return function () {

+       var context = this

+       var args = arguments

+       var later = function () {

+         timeout = null

+         if (!immediate) func.apply(context, args)

+       }

+       var callNow = immediate && !timeout

+       clearTimeout(timeout)

+       timeout = setTimeout(later, wait)

+       if (callNow) func.apply(context, args)

+     }

+   }

+ 

+   function enableSearchInput (enabled) {

+     searchInput.disabled = !enabled

+     searchInput.title = enabled ? '' : 'Loading index...'

+   }

+ 

+   function initSearch (lunr, data) {

+     var start = performance.now()

+     var index = Object.assign({ index: lunr.Index.load(data.index), store: data.store })

+     enableSearchInput(true)

+     searchInput.dispatchEvent(

+       new CustomEvent('loadedindex', {

+         detail: {

+           took: performance.now() - start,

+         },

+       })

+     )

+     var debug = 'URLSearchParams' in globalScope && new URLSearchParams(globalScope.location.search).has('lunr-debug')

+     searchInput.addEventListener(

+       'keydown',

+       debounce(function (e) {

+         if (e.key === 'Escape' || e.key === 'Esc') return clearSearchResults(true)

+         try {

+           var query = searchInput.value

+           if (!query) return clearSearchResults()

+           searchIndex(index.index, index.store, searchInput.value)

+         } catch (err) {

+           if (debug) console.debug('Invalid search query: ' + query + ' (' + err.message + ')')

+         }

+       }, 100)

+     )

+     searchInput.addEventListener('click', confineEvent)

+     searchResult.addEventListener('click', confineEvent)

+     document.documentElement.addEventListener('click', clearSearchResults)

+   }

+ 

+   // disable the search input until the index is loaded

+   enableSearchInput(false)

+ 

+   globalScope.initSearch = initSearch

+ })(typeof globalThis !== 'undefined' ? globalThis : window)

@@ -0,0 +1,1 @@ 

+ <!-- Add any HTML code to show at the top of each page — like a banner to a survey, some announcement, etc. Leave this file empty to not show anything-->

Modified user experience and user journey

My first impression of the fedora docs front page is that it is too wordy. Moving further to the other pages, it was the same experience. Humans have a very short reading attention span and the document will definitely be a boring read thus giving users a bad experience. Also, the lack of structure in the information is equally a problem. After thorough study and research, I decided to map the information into relevant groups into project documents, buttons ( about project, teams, partnerships, community, quick docs and get help) call to action button for downloading the Fedora coreOS. This call to action is necessary because it will enable more users to quickly see the coreOS and download it.

I also incorporated pictures to make the front page engaging and interesting. I also iconized the search bar and the language selector and placed them together with the contribute/edit icon and report icon at the top center of the page for users to easily see it as compared to the former page that had some of the icons in words. I also used the Fedora colors more minimally. I also put the Linux 34 and 35 and the server on the header with full details for easy navigation.

My overall suggestion is that the entire project language needs a review, more images are needed, and proper classification of the items. For instance, a different name should be used to differentiate between the engineering team and the mindshare team who both have different functions.

@ladyluchi I don't see a link to the design. Am I missing it?

@ladyluchi I don't see a link to the design. Am I missing it?

@bcotton It is there but here it is again https://pagure.io/Fedora_docs_frontpage_redesign and also a direct link to the Figma file
https://www.figma.com/file/vR5kJjYRs0NbhrRK9GTrnT/Fedora-Docs-page-redesign?node-id=0%3A1

Thanks for the link. I don't see any indication in this pull request that you created a new repo, so I'm not sure I'd have been able to find it.

As for the design, I like the use of the graphics. I agree that the current page is very wordy.

  • Why does Server have featured billing across the top (alongside F35 and F34)? None of the other Editions get that.
  • The contributor documentation is very hard to find. I completely missed it at first look.
  • "Get Help" should include thinks like Ask Fedora. It should explicitly not contain the Packaging Guidelines, which are contributor documentation.

OK. thanks for the feedback. I will make the necessary corrections

@bcotton If I get you correctly, I assumed that developers who access the documentation page will most probably visit the core Fedora documentation which were captured with those three cards. However, what I could add would be to make it a carousel that includes other documentation links where they can access new releases from the-get-go other than having to click on those Icons to find out what to expect. Or what do you think?

I'm not sure I understand what you're suggesting.

I'm not sure I understand what you're suggesting.@bcotton

In addition to restructuring the information architecture, I had 3 things in mind when I redesigned the front page.
1. The usability- achieved by information restructure, reducing the wordiness of the layout, and using smaller font sizes.

  1. The usefulness: I decided to place the Linux 35, 34, and server in the 'header' position to highlight them and increase their visibility to contributors and page visitors because Fedora Linux 35 is the latest version of Linux released on November 2nd, 2021 which was preceded 6 months ago by Linux 34 and i decided to represent them in carousel form(like a slide show for the three of them).
    I don't understand the question on why the server has a billing feature as I do not see where it is on the redesign. I did not add or remove from the information I saw on the old document on the server

  2. Enjoyability: achieved by incorporating pictures to make it interesting for the readers and contributors. I have added the suggestions you made by changing the call to action button to contribute, and also iconized the download packages function.