How to show the modification date of a file in Jekyll? - date

I know that I can specify a variable date in the YAML frontmatter of a file and access its value with {{ page.date }}. This is cumbersome since one easily forgets to change the date when a file is updated. So how can I access a file's modification date?

This is a relatively new plugin that does what you're looking for:
https://github.com/gjtorikian/jekyll-last-modified-at
(found it while searching Google for alternatives to the other answers in this thread)

From Jekyll 1.x there is a page.path that gives you the filename of the post or page being currently processed. Add the following filter (place e.g. in _plugins/myfilters.rb) to get the modification time of a given filename:
module Jekyll
module MyFilters
def file_date(input)
File.mtime(input)
end
end
end
Liquid::Template.register_filter(Jekyll::MyFilters)
So now you can render the date on your posts and pages, e.g., as
{{ page.path | file_date | date_to_string }}

Based on krlmlr's post, I wrote a pre-commit script to update modification time of the date: field in YAML front matter of the modified files.
#!/bin/sh
# Contents of .git/hooks/pre-commit
git diff --cached --name-status | grep "^M" | while read a b; do
cat $b | sed "/---.*/,/---.*/s/^date:.*$/date: $(date -u "+%Y-%m-%d %T %Z")/" > tmp
mv tmp $b
git add $b
done
It is working fine for me. I actually use update: field to indicate the modified time, and date: field to indicate the creation time.

I don't think it's possible to do this without using the YAML frontmatter. Note, however, that specifying date in the frontmatter actually overrides the date given in the post's filename.
So if you have a file 2013-02-19-my-post.md and in its frontmatter you have date: 2013-02-20, the frontmatter takes precedence, and the post's date is 2013-02-20.
If you want to keep track of when a post is modified, I'm afraid there's no way to do this without using a custom field in your frontmatter. Call it modified-date or something. Yes, it is a bit tedious. I have a TextExpander snippet that automatically outputs current date/time in the proper format, which is handy.

If you're using Git, you can install a pre-commit hook that automatically updates the date field in the front matter of modified files. I was very surprised to learn that a pre-commit hook can indeed change the contents of the commit! You only have to manually install a pre-commit hook into each clone of the repository -- the following rough draft works for me for updating the overall modification time of the entire website:
#!/bin/sh
# Contents of .git/hooks/pre-commit
set -e
set -x
sed -i "s/^date:.*$/date: $(TZ=UTC date "+%Y-%m-%d %H:%M:%S %Z")/" _config.yml
git add _config.yml
The _config.yml is expected to have a field "date", its contents will be replaced by each commit. It can be then accessed via {{ site.date }}. Example:
date: 2015-04-10 10:51:37 UTC
The script can be generalized to update only those pages or posts that have changed (query via git diff-index --cached) -- I don't have the time to implement this but I'd be interested to learn if a neater solution pops out eventually.

Related

Using tac on most recent log file out of several log files in a directory

I have several log files in a directory that we’ll call path/to/directory that are in the following format after long listing in Red Hat Enterprise 6:
-rw-r——-. 1 root root 17096 Sep 30 11:00 logfile_YYYYDDMM_HHMMSS.log
There are several of these log files that are generated everyday. I need to automatically tac the most recently-modified file without typing the exact name of the log file. For example, I’d like to do:
tac /path/to/directory/logile*.log | grep -m 1 keyword
And have it automatically tac the most recently modified file and grep the keyword in the reverse direction from the end of the log file so it runs quicker. Is this possible?
The problem I’m running into is that there is always more than one log file in the /path/to/directory and I can’t get Linux to automatically tac the most recently modified file as of yet. Any help would be greatly appreciated.
I’ve tried:
tac /path/to/directory/logfile_$(date +%Y%m%d)*.log
which will tac a file created on the present date but the part that I’m having trouble with is using tac on the newest file (by YYYYMMDD AND HHMMSS) because multiple files can be generated on the same date but only one of them can be the most current and the most current log file is the only one I care about. I can’t use a symbolic link either.. Limitations, sigh.
The problem you seem to be expressing in your question isn't so much about tac, but rather .. how to select the most recent of a set of predictably named files in a directory.
If your filenames really are in the format logfile_YYYYDDMM_HHMMSS.log, then they will sort lexically without the need for an innate understanding of dates. Thus, if your shell is bash, you might:
shopt -s nullglob
for x in /path/to/logfile_*.log; do
[[ "$x" > "$file" ]] && file="$x"
done
The nullglob option tells bash to expand a glob matching no files as a null rather than as a literal string. Following the code above, you might want to test for the existence of $hit before feeding it to tac.

GitHub URL for latest release of the _download file_?

Although this question is similar to GitHub latest release, it's actually different -- it's about a link that means "the latest version of the download file itself".
GitHub provides a "Latest" URL that redirects to the information page for the latest release. For example: https://github.com/reactiveui/ReactiveUI/releases/latest will redirect to https://github.com/reactiveui/ReactiveUI/releases/tag/5.99.6 (as I type this; or to the page for a newer version, someday).
That's great but I need a URL to the download file itself. In this example, the .zip file associated with the green download button, https://github.com/reactiveui/ReactiveUI/releases/download/5.99.6/ReactiveUI-5.99.6.zip (as I type this; or to a newer zip file, someday).
Why? I want to give the URL to curl, as part of a Travis CI script, to download the latest version.
I guessed at a few URLs like /releases/download/latest/file.zip (substituting "latest" for the version part) and /releases/download/file.zip but those 404.
Is there any way to do this -- in the context of a shell script and curl (note: not in a browser page with JS)?
For releases that do not contain the version number or other variable content in their assets' names, you can use a URL of the format:
https://github.com/owner/repository/releases/latest/download/ASSET.ext
As per the docs:
If you'd like to link directly to a download of your latest release asset you can link to /owner/name/releases/latest/download/asset-name.zip.
Here is a way to do it w/o Github if you have a single download in the release:
wget $(curl -s https://api.github.com/repos/USERNAME/REPONAME/releases/latest | grep 'browser_' | cut -d\" -f4)
It is pretty easy (though not pretty), and of course you can swap out wget for another curl call if you want to pipe it to something.
Basically, the curl call nets you a JSON structure, and I'm just using basic shell utilities to extract the URL to the download.
Very interesting, I haven't noticed a "latest" tag in GitHub-releases yet. As i now figured out, they're given away if you're using the "pre-release"-capabilities of GitHubs release-system. But i don't know any way to access binaries via a latest-path.
I would like to suggest you using git (which is available in your travis-vm) to download the latest tag.
Like Julien Renault describes in his blog post, you will be able to checkout the latest tag in the repository like this:
# this step should be optional
git fetch --tags
latestTag=$(git describe --tags `git rev-list --tags --max-count=1`)
git checkout $latestTag
This solution is based on the assumption that the latest tag is also the latest version.
I use this to get the download URLs in PowerShell 5+ (replace ACCOUNT & REPO)
Invoke-RestMethod -uri https://api.github.com/repos/ACCOUNT/REPO/releases/latest | select -ExpandProperty assets | select -expand browser_download_url
Note if they have more than one package this will be a list. If you want to pick a certain one find a unique part of the name i.e. win for Windows and use:
(replace ACCOUNT, REPO & SELECTOR)
Invoke-RestMethod -uri https://api.github.com/repos/ACCOUNT/REPO/releases/latest | select -ExpandProperty assets | ? { $_.name.Contains("SELECTOR")} | select -expand browser_download_url
As a bonus if you assign the above to a variable you can then grab the file and extract it with the following (assuming you assign to $uri):
Invoke-WebRequest $uri -OutFile "release.zip"
Expand-Archive .\release.zip
In PowerShell 6+ this should work on other platforms than Windows.
On windows, only using powershell, this works for me. It can probably be written a lot shorter.
#Downloads latest paket.bootstrapper.exe from github
$urlbase = "https://github.com"
$latestPage="$urlbase/fsprojects/Paket/releases/latest"
Write-Host "Parsing latest release page: $latestPage"
$page=Invoke-Webrequest -uri $latestPage
$latestBootStrapper=($page.Links | Where-Object { $_.href -match "bootstrapper" }).href
$dlurl="$urlbase$latestBootStrapper"
Write-Host "Downloading paket.bootstrapper.exe from $dlurl"
$wc=new-object net.webclient
$wc.UseDefaultCredentials=$true
$wc.Proxy.Credentials=$wc.Credentials
$wc.DownloadFile($dlurl, (join-path (resolve-path ".\") "paket.bootstrapper.exe"))
$repoName = "PowerShell/PowerShell"
$assetPattern = "*-win-x64.msi"
$extractDirectory = "C:\Users\Public\Downloads"
$releasesUri = "https://api.github.com/repos/$repoName/releases/latest"
$asset = (Invoke-WebRequest $releasesUri | ConvertFrom-Json).assets | Where-Object name -like $assetPattern
$downloadUri = $asset.browser_download_url
$extractPath = [System.IO.Path]::Combine($extractDirectory, $asset.name)
Invoke-WebRequest -Uri $downloadUri -Out $extractPath
You can use curl with https://api.github.com. It gives JSON output from which you can easily extract what you need with jq or your favorite json tool.
For example, using the repository in the question:
gituser=reactiveui; repo=ReactiveUI
tag_name=$(curl -sL https://api.github.com/repos/$gituser/$repo/releases/latest | jq -r '.tag_name'); echo $tag_name
# output: "16.3.10"
tarurl=$(curl -sL https://api.github.com/repos/$gituser/$repo/releases/latest | jq -r '.tarball_url'); echo $tarurl
# output: https://api.github.com/repos/reactiveui/ReactiveUI/tarball/16.3.10
zipurl=$(curl -sL https://api.github.com/repos/$gituser/$repo/releases/latest | jq -r '.zipball_url'); echo $zipurl
# output: https://api.github.com/repos/reactiveui/ReactiveUI/zipball/16.3.10
So you could get the download with a nested curl in a one-liner:
curl -OL $(curl -sL https://api.github.com/repos/filesender/filesender/releases/latest | jq -r '.tarball_url')
This will download the file, and save it with the name of its tag_name, but without extension. So you may want to rename it by appending ".tgz" or ".zip", depending on which you downloaded.
Note for Windows users: curl is now installed by default on Windows too, but beware that it must be called as curl.exe. That's because Powershell has an alias stupidly called "curl" which is not the same!
Centos/RHEL
There are 2 options to download using the URL directly.
Via Github API (using CURL and jq package)
Via Github direct (using CURL and sed)
I am listing a demonstration script for each option.
Option 1
#!/bin/bash
# author: fullarray
# Contribution shared on: stackoverflow
# Contribution shared on: github
# date: 06112022
compose_version=$(curl https://api.github.com/repos/docker/compose/releases/latest | jq .name -r)
get_local_os_build=$(uname -s)-$(uname -m)
curl -L https://github.com/docker/compose/releases/download/$compose_version/docker-compose-$get_local_os_build
Option 2
#!/bin/bash
# author: fullarray
# Contribution shared on: stackoverflow
# Contribution shared on: github
# date: 06112022
get_local_os_build=$(uname -s)-$(uname -m)
compose_latest_version=$(curl -L "https://github.com/docker/compose/releases/download/`curl -fsSLI -o /dev/null -w %{url_effective} https://github.com/docker/compose/releases/latest | sed 's#.*tag/##g' && echo`/docker-compose-$get_local_os_build")
If you are fine with cloning the repository first, you can use git tag, which also allows you to sort the tags by version in various ways.
git clone https://github.com/reactiveui/ReactiveUI.git .
LATEST="$(git tag --sort=v:refname | tail -n1)"
git checkout "$LATEST"
This allows for more flexibility, as you can filter the tags you're not interested in with grep, e.g.:
git tag --sort=v:refname | grep -vE '-RC[0-9]+$' | tail -n1
Here's an excerpt from the documentation on git-tag:
Sort based on the key given. Prefix - to sort in descending order of the value. You may use the --sort=<key> option multiple times, in which case the last key becomes the primary key. Also supports version:refname or v:refname (tag names are treated as versions). The version:refname sort order can also be affected by the versionsort.suffix configuration variable. The keys supported are the same as those in git for-each-ref. Sort order defaults to the value configured for the tag.sort variable if it exists, or lexicographic order otherwise. See git-config(1).
If you really don't want to clone the repository, the --sort option also works with git ls-remote. It'll just take a bit more work to get the part you're interested in:
git ls-remote --tags --sort=v:refname https://github.com/reactiveui/ReactiveUI.git | awk -F'/' '{ print $NF }'
This approach doesn't seem to work all too well for the ReactiveUI repository in particular, because their tags are a bit messy, but it's an option.
Please note, that the sorting isn't quite the same as with semantic versioning, but git does allow you to work around most of these cases. As an example mqtt2prometheus has release candidates using the suffix RC1, RC2 etc., but git sorts 0.1.6-RC1 as being newer than 0.1.6. You can tell git that "RC" is a pre-release suffix to make it sort them correctly.
git tag -c 'versionsort.suffix=-RC' --sort=v:refname | tail -n1
Here's an excerpt from the documentation on git-config:
By specifying a single suffix in this variable, any tagname containing that suffix will appear before the corresponding main release. E.g. if the variable is set to "-rc", then all "1.0-rcX" tags will appear before "1.0". If specified multiple times, once per suffix, then the order of suffixes in the configuration will determine the sorting order of tagnames with those suffixes. E.g. if "-pre" appears before "-rc" in the configuration, then all "1.0-preX" tags will be listed before any "1.0-rcX" tags.
You can also sort by the date of the tag using --sort=taggerdate, that might work better in some situations.
As #florianb pointed out, I should use git.
Originally my .travis.yml was something like:
before_install:
- curl -L https://raw.githubusercontent.com/greghendershott/travis-racket/master/install-racket.sh | bash
This would automatically get whatever the latest version is, from the repo.
But someone pointed out to me that GitHub doesn't want people to use raw.github.com for downloads. Instead people should use "releases". So I was a good doob and manually made a release each time. Then my .travis.yml was something like:
before_install:
- curl -L https://github.com/greghendershott/travis-racket/releases/download/v0.6/install-racket.sh | bash
But it's a PITA to make a release each time. Worse, all .travis.yml files need to be updated to point to the newer version of the file.
Instead -- just use git to clone the repo, and use the file within it:
before_install:
- git clone https://github.com/greghendershott/travis-racket.git
- cat travis-racket/install-racket.sh | bash # pipe to bash not sh!

How do I get the raw version of a gist from github?

I need to load a shell script from a raw gist but I can't find a way to get raw URL.
curl -L address-to-raw-gist.sh | bash
And yet there is, look for the raw button (on the top-right of the source code).
The raw URL should look like this:
https://gist.githubusercontent.com/{user}/{gist_hash}/raw/{commit_hash}/{file}
Note: it is possible to get the latest version by omitting the {commit_hash} part, as shown below:
https://gist.githubusercontent.com/{user}/{gist_hash}/raw/{file}
February 2014: the raw url just changed.
See "Gist raw file URI change":
The raw host for all Gist files is changing immediately.
This change was made to further isolate user content from trusted GitHub applications.
The new host is
https://gist.githubusercontent.com.
Existing URIs will redirect to the new host.
Before it was https://gist.github.com/<username>/<gist-id>/raw/...
Now it is https://gist.githubusercontent.com/<username>/<gist-id>/raw/...
For instance:
https://gist.githubusercontent.com/VonC/9184693/raw/30d74d258442c7c65512eafab474568dd706c430/testNewGist
KrisWebDev adds in the comments:
If you want the last version of a Gist document, just remove the <commit>/ from URL
https://gist.githubusercontent.com/VonC/9184693/raw/testNewGist
One can simply use the github api.
https://api.github.com/gists/$GIST_ID
Reference: https://miguelpiedrafita.com/github-gists
Gitlab snippets provide short concise urls, are easy to create and goes well with the command line.
Sample example: Enable bash completion by patching /etc/bash.bashrc
sudo su -
(curl -s https://gitlab.com/snippets/21846/raw && echo) | patch -s /etc/bash.bashrc

CVS command to get brief history of repository

I am using following command to get a brief history of the CVS repository.
cvs -d :pserver:*User*:*Password*#*Repo* rlog -N -d "*StartDate* < *EndDate*" *Module*
This works just fine except for one small problem. It lists all tags created on each file in that repository. I want the tag info, but I only want the tags that are created in the date range specified. How do I change this command to do that.
I don't see a way to do that natively with the rlog command. Faced with this problem, I would write a Perl script to parse the output of the command, correlate the tags to the date range that I want and print them.
Another solution would be to parse the ,v files directly, but I haven't found any robust libraries for doing that. I prefer Perl for that type of task, and the parsing modules don't seem to be very high quality.

CVS: How to get the date when a tag is created?

We have a CVS repository and we create a tag on the active branch whenever a successful build is done.
Is there any way by which I can determine the date when the tag was created?
Looking into the history doesn't helps since it only tells the date-time stamps of the file when it was modified.
Thanks!
You can easily configure CVS to log all tag-related actions.
In the file '$CVSROOT/CVSROOT/taginfo' you can hook up a
pre-tag script like this:
ALL $CVSROOT/CVSROOT/do_tag
If this script returns a non-zero exit value, the tag operation will
be aborted. This allows for syntax checks on the tag names. You can
also use this hook to send emails, whenever a new release has been
tagged. To write a history of all tag operations, you need to do
something like this in your do_tag file:
#!/bin/sh
TAGHISTORY=~cvs/taghistory.log
echo -n "$(date): user $USER, tag " >> $TAGHISTORY
echo "$*" >> $TAGHISTORY
exit 0
If you have the history function enabled, you can execute
the following command:
cvs history -a -T
It will give you some lines like this, giving you date+time, user, module and tagname of each tagging operation:
T 2011-04-02 07:55 +0000 ralph mylib [testtag:A]
For more information check the cvsbook on history