Add pre-commit and CI

A basic pre-commit setup has been added. CI has been implemented with it
to keep a baseline level of quality. Offending files have been fixed as
well.

.github/workflows/qa.yaml

@@ -0,0 +1,13 @@
+name: QA
+on:
+  pull_request:
+  push:
+    branches:
+      - "main"
+jobs:
+  pre-commit:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v4
+    - uses: actions/setup-python@v5
+    - uses: pre-commit/[email protected]

.pre-commit-config.yaml

@@ -0,0 +1,9 @@
+repos:
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v4.4.0
+    hooks:
+      - id: check-yaml
+      - id: trailing-whitespace
+      - id: end-of-file-fixer
+      - id: mixed-line-ending
+        args: ["--fix=lf"]

archetypes/default.md

@@ -3,4 +3,3 @@ title: "{{ replace .Name "-" " " | title }}"
 date: {{ .Date }}
 draft: true
 ---
-

config.yml

@@ -37,4 +37,3 @@ params:
     URL: "https://github.com/zoftko/blog/blob/main/content"
     text: "Suggest changes"
     appendFilePath: true
-

content/posts/handmade-rf-mixer-board-mc1496.md

@@ -18,7 +18,7 @@ with SMD components (this was my first time soldering them).
 
 # Building the board
 Board layout is very important when designing high speed or high frequency circuits. Some basic
-guidelines are: 
+guidelines are:
 
 * Traces need to be kept as short as possible to minimize the effects associated with transmission lines.
 * Component leads should be kept short, or else they will act as inductors. Their parasitic

content/posts/how-to-use-led1642gw.md

@@ -13,7 +13,7 @@ a serial link. This article documents how to achieve such control with the help
 microcontroller.
 
 The end result is the following, a constant fade in/out of LEDs achieved by varying the brightness
-to which the LED1642GW channels are set. 
+to which the LED1642GW channels are set.
 
 *Note that the blinking can't be perceived by the human eye but gets picked up by the camera.*
 
@@ -106,7 +106,7 @@ void TIM2_IRQHandler(void) {
             le_off();
             clock_off();
             // Turn off timer and notify message has been sent
-            TIM2->CR1 &= ~(TIM_CR1_CEN); 
+            TIM2->CR1 &= ~(TIM_CR1_CEN);
             vTaskNotifyGiveFromISR(led1642TaskHandle, NULL);
         }
     }
@@ -142,4 +142,3 @@ other souls.
   than enough.
 
 Finally, the complete [code implementation can be consulted here](https://github.com/crazybolillo/stm32f401xc/blob/main/src/led1642.c).
-

content/posts/lessons-first-side-project.md

@@ -29,7 +29,7 @@ integration.
 I don't like designing UIs, I can appreciate a well designed one, but usually prefer to stick with
 CLIs. This makes designing the front end very tedious. Once I get a general idea it is not too hard
 to code it. I used [Thymeleaf](https://www.thymeleaf.org/), [daisyUI](https://daisyui.com/)
-and [htmx](https://htmx.org/). 
+and [htmx](https://htmx.org/).
 
 With this I was able to get an acceptable (I think) UI, however there was not even a proper
 landing page and I was not really happy with it. This was an itch I was able to sweep under the rug
@@ -84,4 +84,3 @@ have a finished side project with acceptable quality, you really need to keep it
 I may come back in the future to try something similar but with a much simpler workflow. For the
 time being I am happy with the things I learnt. It sounds cheesy but it does seem like after all,
 failure is the quickest path to success.
-

content/posts/simple-am-transmitter-mc1496.md

@@ -4,7 +4,7 @@ date: 2022-08-04T12:31:54-06:00
 tags: ["Electronics", "RF"]
 ---
 The MC1496 is a Balanced Modulator/Demodulator, capable of producing amplitude modulated signals.
-Its internal structure resembles a Gilbert Cell biased by two current sources (one for each leg of the circuit), 
+Its internal structure resembles a Gilbert Cell biased by two current sources (one for each leg of the circuit),
 both of them being biased by a single diode with a 500 Ohm resistor.
 
 {{< figure src="/img/simple-am-transmitter-mc1496/internals.png" title="MC1496 - Internal Schematic" align="center" >}}
@@ -23,7 +23,7 @@ At least three components are required in order to generate an AM signal:
 * Mixer (modulator)
 
 In order for the transmitter to be heard by a commercial receiver, it needs to transmit in the AM commercial band
-(illegal without a license), this band normally goes from 
+(illegal without a license), this band normally goes from
 [530 to 1700 MHz](https://en.wikipedia.org/wiki/Medium_wave#Spectrum_and_channel_allocation).
 I have no regards for the law, and this design does not have enough power nor an antenna good enough (would need to
 be at least 75 m long) to interfere with commercial broadcasts, so a 1 MHz crystal is used for generating the carrier
@@ -43,11 +43,11 @@ desirable.
 
 And from [Crystal Oscillator Circuits](https://www.amazon.com/gp/product/0471874019):
     > It has very good short-term stability because the crystal’s source and load impedance are mostly capacitive rather
-than resistive, which give it a high in-circuit Q. 
+than resistive, which give it a high in-circuit Q.
 The circuit provides a large output signal and simultaneously drives the crystal at a low power level.
 
 The output is taken from the base of the transistor, where there exists a (relatively) big capacitance, so any
-extra stray capacitance introduced by the next stage won't affect the oscillator too much. 
+extra stray capacitance introduced by the next stage won't affect the oscillator too much.
 
 The next stage consists of a common collector buffer, being biased by a constant current source in an attempt to
 provide a high input impedance, trying to isolate the oscillator as much as possible from the next stages.
@@ -73,15 +73,15 @@ overview of the guidelines and the design steps to follow them were:
 3. Modulating pins(1, 4) must be 2.7V below the carrier pins.
 4. The bias pin(5) must be 2.7V below the modulating pins.
 
-This leaves very little headroom, and it's why inductors were used at the output stage. 
+This leaves very little headroom, and it's why inductors were used at the output stage.
 They act as resistances at the operating frequency, providing the gain necessary, but act as a short circuit when
 it comes to DC biasing.
 
 Bias current (I5) was chosen to be 1mA, considering the diode forward voltage to be 0.75 Vdc at ambient
 temperature, the voltage at the bias pin turns out to be 1.25 V.
 
 Knowing the max and min voltage limits, a resistor network was added to provide bias to the two missing stages
-of the circuit. 
+of the circuit.
 
 Audio is inserted into pin 4, where a variable resistor controls the modulation index.
 [This video](https://www.youtube.com/watch?v=38OQub2Vi2Q) explains how varying the bias voltage affects the

content/posts/stm32-bare-metal-usb-implementation.md

@@ -3,16 +3,16 @@ title: "STM32 Bare Metal USB Implementation"
 date: 2023-07-20T23:35:19-06:00
 tags: ["Electronics", "ARM", "STM32", "USB"]
 ---
-Implementing your own bare metal USB stack may provide certain benefits like little to no 
-dependencies, smaller code size, better understanding of internals and suffering. 
+Implementing your own bare metal USB stack may provide certain benefits like little to no
+dependencies, smaller code size, better understanding of internals and suffering.
 
 This article documents the process I went through to manually set up and get the USB
 core of an STM32F401CCU6 to do what I wanted. My goal was to successfully return a device descriptor.
 Once I got that working I had a greater appreciation for TinyUSB and went with it, however this was
 still a ~~painful~~ great learning experience.
 
 It is expected for the reader to already be familiar with USB as a protocol,
-to get up to speed there are great resources like 
+to get up to speed there are great resources like
 [USB Made Simple](https://www.usbmadesimple.co.uk/).
 
 # Implementation
@@ -22,7 +22,7 @@ for most other chips.
 ## Clock setup
 First step is to produce a 48 MHz clock. The board I had uses a 25 MHz crystal which will be used
 as a source. The system clock will also be derived from the PLL and set to 72 MHz. This last step
-is optional. Be warned that **setting the system clock too low will prevent the USB core from correct 
+is optional. Be warned that **setting the system clock too low will prevent the USB core from correct
 functioning**. The exact frequency can be found on your MCU manual, in this case the AHB frequency
 should be higher than 14.2 MHz. Here is the code used for setup:
 
@@ -33,7 +33,7 @@ while ((FLASH->ACR & FLASH_ACR_LATENCY) != FLASH_ACR_LATENCY_2WS) continue;
 
 // Output the clock to MCO1 for verification
 // Set prescaler for APB1 as to not exceed the max frequency recommended (42 MHZ)
-RCC->CFGR |= (0x06 << RCC_CFGR_MCO1PRE_Pos) | (0x03 << RCC_CFGR_MCO1_Pos) | 
+RCC->CFGR |= (0x06 << RCC_CFGR_MCO1PRE_Pos) | (0x03 << RCC_CFGR_MCO1_Pos) |
     (0x04 << RCC_CFGR_PPRE1_Pos);
 
 // Turn on HSE (crystal oscillator) which will be used as the PLL source
@@ -267,7 +267,7 @@ the source code for the driver.
 
 * During implementation, I had a horrible bug which took me a week to solve. The device was successfully
 writing the descriptor, but was not ACK the Status OUT packet which caused transactions to timeout. You
-can read the whole adventure on the [ST community forums](https://community.st.com/t5/stm32-mcu-products/ep0-in-transfer-times-out/td-p/573140). 
+can read the whole adventure on the [ST community forums](https://community.st.com/t5/stm32-mcu-products/ep0-in-transfer-times-out/td-p/573140).
 Hint: I failed to follow the datasheet and was using an RXFIFO which was not big enough.
 
 ## Full code
@@ -306,4 +306,3 @@ and USB protocol itself. At the time it will seem that there is no information a
 will read over and over any resources you can find on implementing USB manually. Over time it gets
 easier and once you are on the other side it does seem obvious. Writing this post now makes it seem
 like it wasn't that bad after all. Hindsight is always 20/20.
-