zegonix/collection
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Compare commits

base: zegonix/collection:f17b9ef0f614270164ff8eb5c60170dd2c445628
Branches Tags
zegonix/collection:main
...
compare: zegonix/collection:main
Branches Tags
zegonix/collection:main

6 Commits
f17b9ef0f6 ... main

Author SHA1 Message Date
root b7b627d1fe look at rss 2026-06-27 11:21:42 +02:00
scbj 068a83b183 (ascii) removed 'extended' ascii 2026-06-25 13:44:56 +02:00
scbj a4848b9972 TODO gdb breakpoints 2026-06-24 16:23:21 +02:00
scbj dfce75c068 added git-introduction written for CT at InES 2026-06-24 11:50:41 +02:00
scbj 7b909ae4da (make) restructured examples 2026-06-24 11:50:18 +02:00
scbj 62360da2e5 (ascii) added toc 2026-06-24 11:49:33 +02:00
9 changed files with 811 additions and 138 deletions
Expand all files Collapse all files
ascii-table.md
+9 -137
View File
@@ -1,9 +1,15 @@
# ASCII table
Ascii table because websites are unusable these days
Ascii table, because websites are unusable these days
## regular ascii table
> In `Bash` one can input any character by pressing `Ctrl+v` (= `^V`)
> and then entering the code of the character.
>
> Example for ASCII in decimal: `^Vnnn`, where 0 <= nnn <= 255
>
> Example for ASCII in hex: `^Vxnn`, where 0 <= nn <= FF
>
> For Unicode do `^Vu<code>`, where 0 <= code <= FFFF
| **dec** | **hex** | **char** | **description** |
|----------:|----------:|:----------|:------------------------------|
@@ -135,137 +141,3 @@ Ascii table because websites are unusable these days
| 125 | `0x7D` | } | right curly bracket |
| 126 | `0x7E` | ~ | tilde |
| 127 | `0x7F` | DEL | delete |
## extended ascii table
| **dec** | **hex** | **char** | **description** |
|----------:|----------:|:----------|:------------------------------|
| 128 | `0x80` | Ç | majuscule c-cedilla |
| 129 | `0x81` | ü | letter u with umlaut or diaeresis , u-umlaut |
| 130 | `0x82` | é | letter e with acute accent or e-acute |
| 131 | `0x83` | â | letter a with circumflex accent or a-circumflex |
| 132 | `0x84` | ä | letter a with umlaut or diaeresis , a-umlaut |
| 133 | `0x85` | à | letter a with grave accent |
| 134 | `0x86` | å | letter a with a ring |
| 135 | `0x87` | ç | minuscule c-cedilla |
| 136 | `0x88` | ê | letter e with circumflex accent or e-circumflex |
| 137 | `0x89` | ë | letter e with umlaut or diaeresis ; e-umlauts |
| 138 | `0x8A` | è | letter e with grave accent |
| 139 | `0x8B` | ï | letter i with umlaut or diaeresis ; i-umlaut |
| 140 | `0x8C` | î | letter i with circumflex accent or i-circumflex |
| 141 | `0x8D` | ì | letter i with grave accent |
| 142 | `0x8E` | Ä | letter a with umlaut or diaeresis ; a-umlaut |
| 143 | `0x8F` | Å | capital letter a with a ring |
| 144 | `0x90` | É | capital letter e with acute accent or e-acute |
| 145 | `0x91` | æ | latin diphthong ae in lowercase |
| 146 | `0x92` | Æ | latin diphthong ae in uppercase |
| 147 | `0x93` | ô | letter o with circumflex accent or o-circumflex |
| 148 | `0x94` | ö | letter o with umlaut or diaeresis ; o-umlaut |
| 149 | `0x95` | ò | letter o with grave accent |
| 150 | `0x96` | û | letter u with circumflex accent or u-circumflex |
| 151 | `0x97` | ù | letter u with grave accent |
| 152 | `0x98` | ÿ | lowercase letter y with diaeresis |
| 153 | `0x99` | Ö | letter o with umlaut or diaeresis ; o-umlaut |
| 154 | `0x9A` | Ü | letter u with umlaut or diaeresis ; u-umlaut |
| 155 | `0x9B` | ø | lowercase slashed zero or empty set |
| 156 | `0x9C` | £ | pound sign ; symbol for the pound sterling |
| 157 | `0x9D` | Ø | uppercase slashed zero or empty set |
| 158 | `0x9E` | × | multiplication sign |
| 159 | `0x9F` | ƒ | function sign ; f with hook sign ; florin sign |
| 160 | `0xA0` | á | lowercase letter a with acute accent or a-acute |
| 161 | `0xA1` | í | lowercase letter i with acute accent or i-acute |
| 162 | `0xA2` | ó | lowercase letter o with acute accent or o-acute |
| 163 | `0xA3` | ú | lowercase letter u with acute accent or u-acute |
| 164 | `0xA4` | ñ | eñe, enie, spanish letter enye, lowercase n with tilde |
| 165 | `0xA5` | Ñ | spanish letter enye, uppercase n with tilde, eñe, enie |
| 166 | `0xA6` | ª | feminine ordinal indicator |
| 167 | `0xA7` | º | masculine ordinal indicator |
| 168 | `0xA8` | ¿ | inverted question marks |
| 169 | `0xA9` | ® | registered trademark symbol |
| 170 | `0xAA` | ¬ | logical negation symbol |
| 171 | `0xAB` | ½ | one half |
| 172 | `0xAC` | ¼ | quarter, one fourth |
| 173 | `0xAD` | ¡ | inverted exclamation marks |
| 174 | `0xAE` | « | angle quotes, guillemets, right-pointing quotation mark |
| 175 | `0xAF` | » | guillemets, angle quotes, left-pointing quotation marks |
| 176 | `0xB0` | ░ | graphic character, low density dotted |
| 177 | `0xB1` | ▒ | graphic character, medium density dotted |
| 178 | `0xB2` | ▓ | graphic character, high density dotted |
| 179 | `0xB3` | │ | box drawing character single vertical line |
| 180 | `0xB4` | ┤ | box drawing character single vertical and left line |
| 181 | `0xB5` | Á | capital letter a with acute accent or a-acute |
| 182 | `0xB6` | Â | letter a with circumflex accent or a-circumflex |
| 183 | `0xB7` | À | letter a with grave accent |
| 184 | `0xB8` | © | copyright symbol |
| 185 | `0xB9` | ╣ | box drawing character double line vertical and left |
| 186 | `0xBA` | ║ | box drawing character double vertical line |
| 187 | `0xBB` | ╗ | box drawing character double line upper right corner |
| 188 | `0xBC` | ╝ | box drawing character double line lower right corner |
| 189 | `0xBD` | ¢ | cent symbol |
| 190 | `0xBE` | ¥ | yen and yuan sign |
| 191 | `0xBF` | ┐ | box drawing character single line upper right corner |
| 192 | `0xC0` | └ | box drawing character single line lower left corner |
| 193 | `0xC1` | ┴ | box drawing character single line horizontal and up |
| 194 | `0xC2` | ┬ | box drawing character single line horizontal down |
| 195 | `0xC3` | ├ | box drawing character single line vertical and right |
| 196 | `0xC4` | ─ | box drawing character single horizontal line |
| 197 | `0xC5` | ┼ | box drawing character single line horizontal vertical |
| 198 | `0xC6` | ã | lowercase letter a with tilde or a-tilde |
| 199 | `0xC7` | Ã | capital letter a with tilde or a-tilde |
| 200 | `0xC8` | ╚ | box drawing character double line lower left corner |
| 201 | `0xC9` | ╔ | box drawing character double line upper left corner |
| 202 | `0xCA` | ╩ | box drawing character double line horizontal and up |
| 203 | `0xCB` | ╦ | box drawing character double line horizontal down |
| 204 | `0xCC` | ╠ | box drawing character double line vertical and right |
| 205 | `0xCD` | ═ | box drawing character double horizontal line |
| 206 | `0xCE` | ╬ | box drawing character double line horizontal vertical |
| 207 | `0xCF` | ¤ | generic currency sign |
| 208 | `0xD0` | ð | lowercase letter eth |
| 209 | `0xD1` | Ð | capital letter eth |
| 210 | `0xD2` | Ê | letter e with circumflex accent or e-circumflex |
| 211 | `0xD3` | Ë | letter e with umlaut or diaeresis, e-umlaut |
| 212 | `0xD4` | È | capital letter e with grave accent |
| 213 | `0xD5` | ı | lowercase dot less i |
| 214 | `0xD6` | Í | capital letter i with acute accent or i-acute |
| 215 | `0xD7` | Î | letter i with circumflex accent or i-circumflex |
| 216 | `0xD8` | Ï | letter i with umlaut or diaeresis ; i-umlaut |
| 217 | `0xD9` | ┘ | box drawing character single line lower right corner |
| 218 | `0xDA` | ┌ | box drawing character single line upper left corner |
| 219 | `0xDB` | █ | block, graphic character |
| 220 | `0xDC` | ▄ | bottom half block |
| 221 | `0xDD` | ¦ | vertical broken bar |
| 222 | `0xDE` | Ì | capital letter i with grave accent |
| 223 | `0xDF` | ▀ | top half block |
| 224 | `0xE0` | Ó | capital letter o with acute accent or o-acute |
| 225 | `0xE1` | ß | letter eszett ; scharfes s or sharp s |
| 226 | `0xE2` | Ô | letter o with circumflex accent or o-circumflex |
| 227 | `0xE3` | Ò | capital letter o with grave accent |
| 228 | `0xE4` | õ | lowercase letter o with tilde or o-tilde |
| 229 | `0xE5` | Õ | capital letter o with tilde or o-tilde |
| 230 | `0xE6` | µ | lowercase letter mu ; micro sign or micron |
| 231 | `0xE7` | þ | lowercase letter thorn |
| 232 | `0xE8` | Þ | capital letter thorn |
| 233 | `0xE9` | Ú | capital letter u with acute accent or u-acute |
| 234 | `0xEA` | Û | letter u with circumflex accent or u-circumflex |
| 235 | `0xEB` | Ù | capital letter u with grave accent |
| 236 | `0xEC` | ý | lowercase letter y with acute accent |
| 237 | `0xED` | Ý | capital letter y with acute accent |
| 238 | `0xEE` | ¯ | macron symbol |
| 239 | `0xEF` | ´ | acute accent |
| 240 | `0xF0` | ≡ | congruence relation symbol |
| 241 | `0xF1` | ± | plus-minus sign |
| 242 | `0xF2` | ‗ | underline or underscore |
| 243 | `0xF3` | ¾ | three quarters, three-fourths |
| 244 | `0xF4` | ¶ | paragraph sign or pilcrow ; end paragraph mark |
| 245 | `0xF5` | § | section sign |
| 246 | `0xF6` | ÷ | the division sign ; obelus |
| 247 | `0xF7` | ¸ | cedilla |
| 248 | `0xF8` | ° | degree symbol |
| 249 | `0xF9` | ¨ | diaresis |
| 250 | `0xFA` | · | interpunct or space dot |
| 251 | `0xFB` | ¹ | superscript one, exponent 1, first power |
| 252 | `0xFC` | ³ | superscript three, exponent 3, cube, third power |
| 253 | `0xFD` | ² | superscript two, exponent 2, square, second power |
| 254 | `0xFE` | ■ | black square |
| 255 | `0xFF` | nbsp | non-breaking space or no-break space |
examples/make/Makefile → examples/make/romanboard/Makefile
+1 -1
View File
@@ -21,7 +21,7 @@ DEVICE := STM32H573II
# include c sources list
AS_SOURCES :=
C_SOURCES :=
include ./src.mk
include ./sources.mk
# compiler commands
PREFIX := arm-none-eabi-
examples/make/flash.jlink → examples/make/romanboard/flash.jlink
View File
examples/make/sources.mk → examples/make/romanboard/sources.mk
View File
gdb.md
+7
View File
@@ -93,6 +93,13 @@ To reset the program one can use the following command
monitor reset halt
```
## breakpoints
To set breakpoints use the command `b`.
<!-- TODO: elaborate on breakpoint (conditions, etc.) -->
## TUI mode
The `GDB` comes with a Terminal User Interface (TUI) which provides view for
git-introduction/Makefile
+39
View File
@@ -0,0 +1,39 @@
SHELL := /usr/bin/env bash
MD_FILES += readme.md
PDF_FILES := $(MD_FILES:.md=.pdf)
VERBOSITY :=
.PHONY: all clean clean_all test pdf repdf remove_pdf debug_pdf gfm
clean_all:: remove_pdf
pdf: $(PDF_FILES)
repdf: remove_pdf pdf
remove_pdf:
@rm -f $(PDF_FILES) || true
debug_pdf: VERBOSITY := --verbose
debug_pdf: $(PDF_FILES)
%.pdf: %.md
pandoc \
--pdf-engine pdflatex \
--listings \
--template ../template.tex \
-f markdown \
-t pdf \
$(VERBOSITY) \
-o $@ \
$<
git-introduction/readme.md
+752
View File
@@ -0,0 +1,752 @@
# Git Gud (at Git)
This document shall serve as both kickstart and quick reference for
beginners.
Though even experienced users may learn something new.
> For a more complete instructions have a look at [**Pro Git**][progit].
> You can access the book either as HTML document, PDF or EPUB.
> Change the websites language for localised versions of the book.
## Table of contents
<!--toc:start-->
- [Git Gud (at Git)](#git-gud-at-git)
- [Table of contents](#table-of-contents)
- [Terminology](#terminology)
- [Git & Github](#git-github)
- [Git help](#git-help)
- [Configuration](#configuration)
- [Creating a repository](#creating-a-repository)
- [The concept of commits](#the-concept-of-commits)
- [Creating a commit](#creating-a-commit)
- [.gitignore files](#gitignore-files)
- [Branches](#branches)
- [Creating/switching a branch](#creatingswitching-a-branch)
- [Merging/rebasing a branch](#mergingrebasing-a-branch)
- [*Merge* or *rebase*?](#merge-or-rebase)
- [Merging](#merging)
- [Rebasing](#rebasing)
- [Resolving conflicts](#resolving-conflicts)
- [History inspection](#history-inspection)
- [History rewriting](#history-rewriting)
- [Troubleshooting](#troubleshooting)
- [I lost a commit!](#i-lost-a-commit)
- [I tried to pull, but got a merge conflict?!](#i-tried-to-pull-but-got-a-merge-conflict)
- [Further reading](#further-reading)
- [Glossary](#glossary)
<!--toc:end-->
## Terminology
Command examples may have arguments enclosed with angled brackets
(`<>`).
That means the user needs to substitute the argument with their
value.
The text inside the angled brackets describes the expected value.
Additionally an argument may be enclosed by brackets (`[]`).
That means the argument is optional.
## Git & Github
Some might be confused about Git and Github
They have similar names, with Github just adding a 'hub'.
* **Git** is a version control system, that lets you save, analyse and
edit the history of files.
It can be used locally, or in combination with a server.
The latter allows for collaboration with a team and can act as backup.
Git is licensed under GPLv2.
* **Github** is a frontend for Git, which adds functions like a
webserver, access control, pull requests or issues to name a few.
It is not a usable service on its own.
Github is proprietary and owned by Microsoft since 2018.
## Git help
When unsure about a Git (sub)command append `--help`.
This opens the most specific manual available for your command.
Example:
```sh
# show help for the 'config' subcommand
git config --help
# show help for the 'remote add' subcommand
git remote add --help
```
## Configuration
Before starting to use Git it is recommended to configure the user
name and email at the very least.
Otherwise, Git will ask you at the first action which requires an
identity, most probably `push`.
Git looks for a file `~/.gitconfig` (`~` = users home directory).
You can either (create and) edit the file directly, or use
`git config`.
See the example below for how to configure the user name and mail.
```
[user]
name = student
email = student@students.zhaw.ch
```
Alternatively use the following commands to let Git write the
configuration file.
```sh
git config set user.name <you-display-name>
git config set user.email <your-address@domain.global-domain>
```
> For Linux users keeping their configuration files in a repository,
> distributing them with symbolic links, the Git configuration can
> include files.
> So the user and email can be stored in a separate file.
>
> ```
> [include]
> path = /path/to/include/file
> ```
Sensible options to set are the editor used for commit messages,
the default branch name and whether to merge or rebase.
The value of editor is the launch command of the desired editor.
Below is an excerpt from my configuration as an example.
```git
[core]
editor = nvim
[init]
defaultBranch = main
[pull]
rebase = true
```
## Creating a repository
When working with a Git frontend like Github there are two ways to
start a repository:
* **Local:** Use `git init` to initialise the current directory as
a Git repository. This will create a directory `.git` which contains
the local information, e.g. history, commit hashes, hooks, etc.
If you want to synchronise a local repository with a server, follow
the steps for creating a remote repository and use `git remote add`
to add a remote and its URL (default remote name is `origin`).
* **Remote:** Open the frontend of choice and create a new repository.
On Github there is usually a green button in the top right corner.
Choose a name, description and licence.
Now get the URL to clone the repository locally, or set it as remote
in an existing one.
On Github click the green `Code` button and choose either **https**
or **ssh** (recommended, but requires additional setup).
See examples for the Git commands mentioned in this section below.
```sh
# initialise directory as git repository
# use '.' as path to initialise current directory
git init [<path/to/directory>]
# add remote to local repository under name 'origin'
git remote add origin <url-to-remote-repository>
# change URL, for example to switch to ssh after cloning
# a repository over https
git remote set-url <remote-name> <remote-url>
# clone a remote repository
git clone <repository-url>
```
> Mind, that a remote is not strictly necessary, because the local
> repository contains the full information for version control.
> A remote is useful as a backup of your information and to work with
> other contributors.
> Choosing a licence is important, if only to deny liability for
> your code.
> I recommend to setup a pair of ssh keys and use ssh connections
> to interact with the Git server.
> It is more secure and usually more convenient too.
## The concept of commits
Before looking at how to add or change commits, it is prudent to get
an understanding of what a commit is.
A Git history is a linked list of changes, called commits.
Each commit has one (or in case of merge commits more) parents and an
arbitrary number of children.
A commit is a delta relative to its parent(s) and does **not**
directly contain the current state.
## Creating a commit
Use `git add <file-or-directory>` to stage changes made to a file
or directory (a directory means all contained files, recursively).
List the changed files in your working tree with `git status` and show
the differences with `git diff [<path>]`.
With `git commit` you can commit the staged changes.
This will open the configured editor.
Enter your commit message at the very top, save and close the file.
```
# Please enter the commit message for your changes. Lines starting
# with '#' will be ignored, and an empty message aborts the commit.
#
# On branch main
# Your branch is up to date with 'origin/main'.
#
# Changes to be committed:
# modified: git-introduction/readme.md
#
```
You now created a commit, but it is only local, use `git push` to
upload it to the remote repository
(This may fail due to conflicts, see
[Resolving conflicts][conflicts]).
Instead of opening an editor you can also commit with
`git commit -m <commit-message>` but I prefer opening an editor,
because it gives me another chance to spot missing changes, or changes
I did not want to add to the commit.
> It is sensible to make small commits with coherent scope.
>
> When adding a source file it is reasonable to commit the change to
> the Makefile too.
>
> However committing two code changes for distinct functionality at
> once is frown upon.
>
> The reason is, that when searching the history for a commit, it gets
> really difficult to find the change you are looking for, when
> commits contain a lot of changes.
## .gitignore files
Committing files generated by the project is a **crime**, punishable
with a sentence to read the entire Git documentation.
Manually ignoring the generated files would be quite tedious, thus
Git has a mechanism to automatically ignore changes matching a user
defined pattern.
It is very simple, just add a file `.gitignore` to any directory in
your repository.
Such an ignore file is applied for the directory they reside in and
all lower levels.
An ignore file has no effect on its parent or sibling directories.
Here is an example of a `.gitignore`:
```git
# ignore build directory (and all its contents)
build/
# ignore all pdf
*.pdf
# do not ignore specific pdf
!readme.pdf
```
## Branches
A branch is essentially a named reference to a list of commits,
originating from a commit on the main branch.
Here is an ASCII graphic to try and explain the concept.
Keep in mind, that commits (represented by capital letters) contain
only the differences to their parents.
```
A -- B -- D ------ G (main)
\
C -- E -- F (branch)
```
Branches are a good tool to work on a project with several
contributors without solving conflicts on every commit.
Or to quickly try an idea without the risk to poison the development
branch.
Imagine the a scenario, where you work with a colleague on a project.
Your colleague works one part and you on another but both of you have
to make changes to the glue logic inbetween.
The one who commits second will have to solve conflicts.
This still applies when both of you work on separate branches, but
only once, when you rebase or merge your branch back to mainline.
While working on your branch you never need to solve conflicts.
Another use for branches is to develop several features concurrently.
Working with branches you jump between the features without having to
worry about one to interfere with another.
Also you can have a branch reviewed while working on a different
feature.
### Creating/switching a branch
There are at least two methods to create a branch.
Here are the two that come to mind first:
```
# create a branch,
# but do not switch to the created branch
git branch <branch-name>
# create a branch and switch to it
git checkout -b <branch-name>
```
If your working tree is clean, you can switch between branches with
the following commands:
```
# change to branch (or commit)
# (does not create branches or commits)
git checkout <branch-name/commit-hash>
# change to branch
git switch <branch-name>
```
## Merging/rebasing a branch
When a feature is implemented and tested, it is usually to be
integrated into the main branch.
Git provides two ways to achieve this, merging and rebasing.
The following subchapter explains the two philosophies.
### *Merge* or *rebase*?
> My quick answer is: Does not matter, but use only one strategy per
> repository.
This chapter will give a quick explanation of what *merging* and
*rebasing* means and at the end I explain my opinion.
**Merging** means creating a commit with two parents.
When merging, all conflicts are resolved in one merge commit.
Here is some ASCII art to visualise this concept:
```
# merge
A -- C (main)
\
B -- D -- E (feature branch)
| | | |
v v v v
A -- C ------ F (main)
\ /
B -- D -- E (feature branch)
```
**Rebasing** means to recalculate the deltas from a new origin.
This way one can just add (fast forward) the commits of the feature
branch on top of the main branch.
When rebasing, conflicts are resolved in the commit of the feature
branch, that introduces the conflict.
Here is some ASCII art to visualise this concept:
```
# rebase
A -- C (main)
\
B -- D -- E (feature branch)
| | | |
v v v v
A -- C (main)
\
B' -- D' -- E' (feature branch)
# rebased feature-branch on top of main
| | | |
v v v v
A -- C -- B' -- D' -- E' (main)
# rebased main on top of rebased feature-branch
# (no conflicts -> fast forward)
```
**Conclusion** both strategies have advantages and disadvantages:
* *Merging* has the advantage of keeping more information.
* *Rebasing* has the advantage of better readability of the history.
* Some actions for history cleanup require the interactive *rebase*
tool, which may be a *rebasing* argument for some.
I personally have a preference for *rebasing* due to the cleaner
history, but find it perfectly reasonable to use *merging*.
### Merging
When merging, checkout the destination branch and merge the feature
branch into it.
You will get one merge commit, that solves all conflicts.
Both branches stay in your history and are displayed side to side in
the graph view.
Here I explain the procedure of a merge, below you find the commands
used.
1) First synchronise the local with the remote repository.
1) Checkout the **destination** branch.
1) Merge the feature branch into your current (destination) branch.
1) This may cause conflicts.
1) Get a list of files containing conflicts with `git status`.
1) Resolve conflicts (see [Resolving conflicts][conflicts])
1) Continue with `git merge --contine`, choose a message for the
merge commit, save and close.
1) Synchronise the remote with `git push`.
### Rebasing
When rebasing, checkout the feature branch and rebase it on the main
branch.
This way all conflicts are resolved on the feature branch and the
commits can be fast forwarded to the main branch.
Rebasing requires a force push.
Many projects prohibit force pushes on the main branch, so rebasing
the main branch on top of the feature branch is not even an option.
> Be very careful with force pushes.
>
> Always substitue `--force` with `--force-with-lease`.
> This at least disables one to overwrite commits on the remote,
> that are not known locally.
Here I explain the procedure of a rebase, below you find the commands
used.
1) First synchronise the local with the remote repository.
1) Checkout the **feature** branch.
1) Rebase the feature branch to the destination branch
1) This might cause conflicts.
1) Get a list of files containing conflicts with `git status`.
1) Resolve conflicts (see [Resolving conflicts][conflicts]).
1) Continue the rebase with `git rebase --continue`.
1) Repeat until all conflicts are solved.
When rebasing, conflicts are solved in the commit, which introduces
them.
Thus one may have to repeat these steps as many times, as there are
commits to rebase.
1) Force push your rebased branch (with `--force-with-lease`).
1) Check out the destination branch.
1) Rebase destination branch to feature-branch.
1) Push rebased destination branch (this does not require force).
```sh
# fetch the newest state from the remote
git fetch --all
# switch branch
git checkout <feature-branch>
# rebase current branch onto <branch>
git rebase <branch>
# show files containing conflicts during a rebase
git status
# stage solved conflicts
git add <file/directory>
# continue rebase after conflicts are solved
git rebase --continue
# push rebased commits to feature branch
#
# NEVER USE `--force`, SUBSTITUE WITH `--force-with-lease`!
#
# be very careful with force pushes `--force-with-lease` is
# still a potential foot gun!
git push --force-with-lease
```
## Resolving conflicts
For resolving conflicts you can either use a graphical mergetool, or
decide to do it manually (in the file itself).
I do not use any GUI for Git, so can not give you any
recommendations, or instructions for how to use them.
I will just describe the manual approach, it is no more difficult
really.
Git will insert searchable sequences around the conflicts and label
both versions with the source of the change.
See the example below.
```
<<<<<<< HEAD
hello world
=======
blargh
>>>>>>> f837762 (b)
```
To solve a conflict, decide which parts you want to keep and delete
the rest of the changes encased by Gits markers.
If you want or need to you can also mix the changes of both commits.
Then delete the markers and save the file.
Stage the file and continue the merge or rebase with
`git merge --continue` or `git rebase --continue` respectively.
## History inspection
You can inspect the history with `git log`.
Use `-n <number>` or just `-<number>` to show only <number> entries.
Use `--all` to show commits newer than HEAD, other branches and
remote references.
Use `--oneline` to get a compact view. Combine it with `--graph` to
display the commit tree as ASCII art.
Use `git show <commit-hash/commit-reference>` to get the diff of the
commit.
## History rewriting
This is an advanced topic not recommended for beginners.
I suggest to get used to work with Git first and come back once the
desire to clean up the history rises.
The primary tool for history rewriting is the interactive rebase
(`git rebase -i`).
It provides a multitude of functions, reorder, reword and squash to
name just a few.
Using `git rebase -i <commit-reference/commit-hash>` will open a list
of all commits following the provided commit.
In the development repository for CT for example I used the command
`git rebase -i HEAD~3` to get the following file opened by Git.
*The '~3' means three commits before HEAD.*
```
pick a958dff # continued ct2->lab 05
pick 52c15fd # elaborated root readme for ct1
pick 555cf90 # moved and extended `git-introduction.md`
# Rebase 4abb7c9..555cf90 onto 4abb7c9 (3 commands)
#
# Commands:
# p, pick <commit> = use commit
# r, reword <commit> = use commit, but edit the commit message
# e, edit <commit> = use commit, but stop for amending
# s, squash <commit> = use commit, but meld into previous
# commit
# f, fixup [-C | -c] <commit> = like "squash" but keep only the
# previous commit's log message, unless -C
# is used, in which case keep only this
# commit's message; -c is same as -C but
# opens the editor
# x, exec <command> = run command (the rest of the line) using
# shell
# b, break = stop here (continue rebase later with 'git rebase
# --continue')
# d, drop <commit> = remove commit
# l, label <label> = label current HEAD with a name
# t, reset <label> = reset HEAD to a label
# m, merge [-C <commit> | -c <commit>] <label> [# <oneline>]
# create a merge commit using the original
# merge commit's message (or the oneline, if
# no original merge commit was specified);
# use -c <commit> to reword the commit
# message
# u, update-ref <ref> = track a placeholder for the <ref> to
# be updated to this position in the new
# commits. The <ref> is updated at the end
# of the rebase.
#
# These lines can be re-ordered; they are executed from top
# to bottom.
#
# If you remove a line here THAT COMMIT WILL BE LOST.
#
# However, if you remove everything, the rebase will be
# aborted.
```
The commits are listed oldest to youngest from the top.
I could reorder them by just moving the lines.
Or remove a commit by changing the 'pick' at start of its line to
'drop'.
I will not go into detail on the available actions, just one note on
'squash' (combine), which is a function I use regularly.
Commits are always squashed into their parents (upwards in this list).
After editing the list as intended, save and close the file.
If you chose actions, which need further input, Git will
open another file.
If you for example chose to reword a commit, Git will open a file with
the old commit message, which you can edit, save and close.
Many functions require a force push.
**Again be very careful with force pushes**.
<!-- TODO:
## General rules
**Do**
* commit small, cohesive changes
* commit regularly
* create a new, well named branch for every feature
* merge branches promply (do not let them sit for too long)
* add a `.gitignore` with rules sensible for the project
**Do not**
* commit files generated from the repositories sources, files
generated elsewhere, which do not change, may be fine
* force push unless it is your feature branch (there are **rare**
exceptions for this rule)
*
-->
## Troubleshooting
### I lost a commit!
Do not worry, use `git reflog`.
The reference log keeps track of references and usually remembers
"dangling" commits.
This allows you to checkout or cherry-pick a "lost" commit or
reference.
### I tried to pull, but got a merge conflict?!
That is why you should always fetch first and pull only, after
checking for conflicts.
This problem can occur, when someone rewrote the history and your
local HEAD points to a commit that does no longer exist.
The solution is very easy:
`git reset --hard <remote-name>/<branch-or-reference-name>`
If you have local commits that conflict with commits on the remote,
see [Merging](#merging) or [Rebasing](#rebasing).
## Further reading
Here a few functions this guide does not cover, for you to read up
on, if interested:
* `git stash` allows to stash local changes, enabling actions like
pull or rebase.
* `git blame` can determine, who is responsible for code changes
(line by line).
* `git cherry-pick` allows to add single (or several) commits from a
feature branch to the current branch, without having to add the
entire feature branch.
* `git bisect` uses binary search to determine the commit, which
introduced a bug.
* **command aliases** let you access often used command (with flags)
via `git <alias>`.
(This is distinct from the shells aliases.)
## Glossary
+-------------------+-----------------------------------------------+
| **Term** | **Description** |
+:==================+:==============================================+
| branch | Separate named reference originating from an |
| | older commit on the source branch. |
| | A branch can originate from main/master |
| | (the trunk in the tree analogy) or any other |
| | branch. |
+-------------------+-----------------------------------------------+
| commit | Repository state defined as delta to its |
| | parent. |
| | Each commit is identified by its commit hash, |
| | a 40 byte long string of alphanumeric |
| | characters. |
| | The hash can be used in many Git commands, |
| | e.g. `diff`, `reset` or `checkout`. |
| | A commit usually has one parent, but can have |
| | an arbitrary number of children, one for each |
| | branch originating from the commit. |
+-------------------+-----------------------------------------------+
| commit reference | Named pointer to a specific commit. |
| | An example would be `HEAD`, which points to |
| | the currently checked out commit. |
| | Branches and tags are references too. |
+-------------------+-----------------------------------------------+
| head | Pointer to commit currently checked out. |
| | On remote it will usually point to *main*. |
| | With `git checkout` one can checkout a |
| | to e.g. test something or make new branch. |
| | In such a case the head would point to that |
| | commit. |
+-------------------+-----------------------------------------------+
| index / | `git add` puts file/directories (rather the |
| staging area | changes made to them) into a cache, called |
| | *index* or *staging area* (in arguments it is |
| | also called chache). |
| | `git commit` adds the changes in that cache |
| | to the history with a hash, a descriptoin and |
| | a few more information about the commit. |
+-------------------+-----------------------------------------------+
| working tree | Local copy of the repository. |
| | a working tree is clean if it has no |
| | uncommitted changes. |
+-------------------+-----------------------------------------------+
<!-- links -->
[conflicts]: #resolving-conflicts
[progit]: https://git-scm.com/book/en/v2
git-introduction/resources/example-commit.png
BIN
View File
Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 58 KiB

readme.md
+3
View File
@@ -5,6 +5,7 @@ This repository holds information I wrote down for quick access.
# Interesting topics to read up on
* [RSS feeds][ls-rss-feeds] & [newsraft][newsraft]
* [LaTeX package *tikz*][tikz]
* [HSV/HSL and colourspaces in general][hsl]
* [GCCs handling of large immediate values][gcc-literals]
@@ -12,6 +13,8 @@ This repository holds information I wrote down for quick access.
<!-- links -->
[ls-rss-feeds]: https://www.youtube.com/watch?v=hMH9w6pyzvU&list=PL-p5XmQHB_JRRnoQyjOfioJdDmu87DIJc&index=4
[newsraft]: https://codeberg.org/newsraft/newsraft/releases
[tikz]: https://nathantypanski.com/blog/2014-10-29-tikz-timing.html
[hsl]: https://en.wikipedia.org/wiki/HSL_and_HSV#Use_in_end-user_software
[gcc-literals]: https://medium.com/@python-javascript-php-html-css/understanding-gccs-handling-of-large-immediate-values-in-armv7-assembly-d8a906231b44