* Introduce calibration settings for all touchscreen drivers.
this will override the common values.
The x,y coordinates only calculated when the right calibrations are set.
* resolve issues reported by CI
* remove unneeded spaces and newlines
* Forgot to remove some obsolete code
* remove get_setup_priority from xpt2046
* remove media_player changes.
* media_player: removed to much,
* Update suggestions
* referd back the `get_setup_priority` removal so it can be moved into a othe PR.
* tt21100: restore init read
* fix spacing
* load native display dimensions instead of using internal dimensons.
and load it only onse on setup
* moved `update_touches()` to protexted section
* adding Clydes PR#6049
* add multitouch test script
* Update all Touchscreen replacing `get_*_internal` to `get_native_*`
* fixed some CI recomendations
* couple of fixes
* make sure the display is running before touchscreen is setup
* fix clang
* revert back last changes
* xpt2046: change log level for testing
* logging information
* add test file
* fix polling issue with the for example the xpt2046
* fixed some CI issues
* fixed some CI issues
* restore mirror parameter discriptions
* same for the swap_xy
* same for the transform
* remove the above const from const.py
* and put the above const bacl const.py
* Merge branch 'nvds-touchscreen-fix1' of https://github.com/nielsnl68/esphome into nvds-touchscreen-fix1
* and put the above const bacl const.py
* [tt21100] making interupt pin optional
* [tt21100] making interupt pin optional (now complete)
* update the display part based on @clyde'
s changes.
* fix issue with ft6x36 touvhscreen
* reverd back touch check. add comment
* add some extra checks to the ft6x36
* add an other log and a typo fixed
* okay an other fix.
* add an extra check like others do
and fix data type
* [ft6336] fix update race when ts is touched.
* [touchscreen] update some log's with a verbose level.
* fix clang issues
* fix the clang issues
* fix the clang issues
* fix virtual issue.
* fix the clang issues
* an other clang issues
* remove anti-aliased fonts support. It does not belong here.
* remove anti-aliased fonts support. It does not belong here.
* rename test script
* Moving the test files to there right location.
* rename the test files
* clean up the code
* add a new line
* clang fixings
* clang fixings
* remove comment
* remove comment
* Update esphome/components/touchscreen/__init__.py
Co-authored-by: guillempages <guillempages@users.noreply.github.com>
* Update esphome/components/touchscreen/__init__.py
Co-authored-by: guillempages <guillempages@users.noreply.github.com>
* Update esphome/components/touchscreen/__init__.py
Co-authored-by: guillempages <guillempages@users.noreply.github.com>
* Update esphome/components/touchscreen/touchscreen.cpp
* Update esphome/components/touchscreen/touchscreen.cpp
* [ft63x6] add threshold
---------
Co-authored-by: Clyde Stubbs <2366188+clydebarrow@users.noreply.github.com>
Co-authored-by: guillempages <guillempages@users.noreply.github.com>
* Nextion TFT upload IDF memory optimization
This optimizes the memory in use for TFT upload when using `esp-idf` framework.
Basically, the engine establishes 3 connections to the the http/https server:
1. Fetch the file size (used to manage chunks and file size)
2. Transfer the 1st chunk (when it evaluates Nextion response to define either to continue from that point or to another point in the file)
3. Transfer the remaining data.
Until now, connection 1 was kept open during the whole process taking aprox 40kb of heap in a esp32dev (NSPanel in my tests) and the same amount of memory was needed to the 2nd and 3rd connections (which never competes to each other).
With this change, each connection is closed and released before opening the next one with a significant reduction on the required heap needed for this transfer.
This can still be improved to use a persistent connection, but I will look at this in the future, so it is not part of this change.
In addition to the better connection management, I've added quite a lot of log (mostly at VERBOSE level), which was used for troubleshooting here.
I was unsure about removing this. As it can be useful for others, I decided to keep it, but I will be fine about removing it if this is now in line with ESPHome best practices.
* clang-format
* Log response length
DAVe3283