The boundary parsing in the webserver could end up missing boundaries if the
uploaded file had `--` at the start of the line because it read in the entire boundary
length worth of bytes. Fix by only reading up to either the boundary length or
a newline, avoiding the issue.
Fixes#7542
The SDFS implementation didn't include plumbing to support user-supplied
timestamp generators (for file create/write times) because the SDFat
library doesn't support a callback parameter.
Work around it by using a single, global (inside sdfs namespace) that
the static timestamp callback member can see and use.
Add a test of the feature in CI.
Fixes#7682
Fixes#7671
Allows for code to do things like read certs from LittleFS or even HTTP
connections with code like:
File cert = LittleFS.open("/client-crt.pem", "r");
clientCert = new X509List(cert, cert.size());
cert.close();
* Add Print::availableForWrite method
Adds an availableForWrite() method to the Print class, matching current
ArduinoCore-API commit 398e70f188e2b861c10d9ffe5e2bfcb6a4a4f489 .
Hook availableForWrite into the SDFS filesystem (other FSes don't have
this capability built-in).
Fixes#7650
* WiFiClient::availableForWrite proto matching Print
* Fix Netdump signedness warning
* Clean up Serial availableForWrite
This is evidently a breaking change due to the type difference.
Arduino's `availableForWrite` returns an `int`, while the
(multiply-implemented, non-virtual) core `availableForWrite` returned
`size_t`.
When building with the Netdump library, the IDE generates the following warning:
> WARNING: library Netdump claims to run on (esp8266 lwip) architecture(s) and may be incompatible with your current board which runs on (esp8266) architecture(s).
Fix the library.properties to call the architecture just "esp8266"
* Add Wrong Password wifi status case
* Add wrong password case for status return
* Add wrong password case for debug
* Add Wrong password case to interactive example
* Add case for wrong password to station doc
* Add case for wrong password to resumeFromShutdown
* Add wrong password case to wifi readme
* Update ESP8266WiFiGeneric.cpp
Fixed critical issues WiFiMulti library:
- WiFi scan timeout implemented to prevent endless connect loop
- Fallback implemented on WiFi connect failure to prevent endless loop
- Fast WiFi connection at startup
- Improved debug prints
- Doxygen added
- Code maturing
- Example update
Make functions not related to ESP8266WiFiMulti class static
Revert static functions startScan and printWiFiScan()
Use PolledTimeout.h to protect while loops
Move static functions beginning of the file
Add connect timeout to example
* Update certs-from-mozilla.py
Check if ar exists, if not tell the user what to get to prevent issue #7300 also dynamically get certs instead of hardcoded row item https://github.com/esp8266/Arduino/pull/7573#issuecomment-686192353
changed comment for missing ar exception
updated path and check for openssl
Simple example update to pass the method as a parameter to getDigestAuth(), so it is more easily used for POST.
Add setting the ransom seed to RANDOM_REG32 in setup() for better getCNonce() values.
The logic can be simplified by using integer logic without a functional
change. Reduces code size by 40% (78 bytes -> 46 bytes) and silences
a Warith-conversion warning.
Converting floats to doubles is very expensive on esp8266, so prefer
calculations or comparisons as float. This saves 10% (20 bytes) of the
String::parseFloat() code size and probably quite a bit of runtime
overhead.
* Clean up minor warnings from LGTM.com
LGTM (Semmie) is a tool, bought by GitHub last year, that conducts basic
linting tasks on code and HTML.
Clean up the warnings identified in the latest report:
https://lgtm.com/projects/g/esp8266/Arduino/?mode=list
No functionality should change, however this may fix some issues with
the perl utilities not exiting properly on a Ctrl-C from the command
line.
* Back out HTML changes and rerun boards.txt.py
Run valgrind on host mock example runs to catch more bugs in CI. These
tests would have caught the problem in #7464 before users did.
Add a list of some randomly picked examples to run, and add an option to
run the loop exactly once in the host mock routine, so the test will
actually exit under valgrind.
