How to properly include an http C library into an Android Project? - c

I am making an Android Project and I am trying to make a post request using HTTPS to a server. To make the post request I am requires to use JNI so I need to implement this in C.
My idea was to use a library that I could include into my project as i have done with the minizip library here.
I found this library here that seemed to me light engough and could server my purpose. I included in the folder c as you can also see below the files ca_cert.h, https.c and https.h along with the folder mbedtls exactly as is found on the github project.
├── app.iml
├── build.gradle
├── CMakeLists.txt
└── src
├── androidTest
│   └── java
│   └── etc
├── main
│   ├── AndroidManifest.xml
│   ├── c
│   │   ├── ca_cert.h
│   │   ├── https.c
│   │   ├── https.h
│   │   ├── mbedtls
│   │   │   ├── etc
│   │   ├── native-lib.c
│   │   ├── pathfinder.c
│   │   ├── pathfinder.h
│   │   ├── post_data.c
│   │   ├── post_data.h
│   │   ├── third
│   ├── java
│   │   └── etc
│   └── res
│   ├── etc
└── test
└── java
└── etc
As you can see in the tree structure above I have a CMakeLists.txt in the root directory and in src/main/c/ you will see the files and the folder i took from the the https library mentioned.
The contents of the CMakeLists.txt are the following
cmake_minimum_required(VERSION 3.4.1)
add_library( # Sets the name of the library.
native-lib
# Sets the library as a shared library.
SHARED
# Provides a relative path to your source file(s).
src/main/c/native-lib.c
src/main/c/ca_cert.h
src/main/c/https.h
)
include_directories(native-lib SHARED src/main/c/mbedtls/include)
#add_definitions(-DHAVE_ZLIB)
find_library( PRE_BUILD_ANDROID_LIB android)
find_library( log-lib log)
find_library(z-lib z)
target_link_libraries( native-lib
${PRE_BUILD_ANDROID_LIB}
${log-lib}
${z-lib}
)
I am missing something though for sure cause when i attempt to run a simple example like the following
JNIEXPORT jint JNICALL
Java_example_example_do_post(
JNIEnv *env,
jobject this ) {
NSV_LOGE("post_data starts\n");
char *url;
char data[1024], response[4096];
int i, ret, size;
HTTP_INFO hi1, hi2;
// Init http session. verify: check the server CA cert.
http_init(&hi1, FALSE);
http_init(&hi2, TRUE);
url = "http://httpbin.org/get?message=https_client";
ret = http_get(&hi1, url, response, sizeof(response));
return 0;
}
I get
../../../../src/main/c/native-lib.c:56: error: undefined reference to 'http_init'
../../../../src/main/c/native-lib.c:57: error: undefined reference to 'http_init'
../../../../src/main/c/native-lib.c:61: error: undefined reference to 'http_get'
clang: error: linker command failed with exit code 1 (use -v to see invocation)

Replace
src/main/c/ca_cert.h
src/main/c/https.h
with
src/main/c/https.c
in CMakeLists.txt.

Related

add triggers manually inside my migrations with prisma ORM

I am trying to create triggers using postgresql with prisma as ORM
but it seems that it haven't supported triggers yet.
and I found this answer ,but i don't want to create my triggers logic in the server level (using middlewares).
now I am asking, can i make my triggers manually inside my database migrations as a migration level, or it may cause some problems?
for example i have the following migrations inside my application
├── prisma
│   ├── migrations
│   │   ├── 20230202011931_initial
│   │   │   └── migration.sql
│   │   ├── 20230202012555_default_to_created_at
│   │   │   └── migration.sql
│   │   ├── 20230202130457_mapping
│   │   │   └── migration.sql
│   │   └── migration_lock.toml
│   └── schema.prisma
can i choose one of these files to add my triggers inside or just create another file and add my triggers inside it?
Yes, you can declare database triggers that are a part of migrations.
You can create a migration with --create-only flag.
So invoking the command npx prisma migrate dev --create-only will create an empty migration file. You can then add the database trigger in this newly created file. Invoking npx prisma migrate dev again would then apply this new migration.
You can read more about customizing migrations here.

CMake import external library (GNU readline) using ExternalProject_Add()

I am really new (just a few weeks) to C and its bulding tools. Using the CMake documetation and some other guides I created a basic CMake project. This works as expectd.
Now I wanted to use some libraries in my project (GNU readline and cJSON).
The cJSON library also uses CMake so I managed to include it in my project without too much trouble.
The GNU readline library on the other hand uses ninja and autotools (if I'm not mistaken). This is where I'm facing issues.
I am trying to build the GNU readline library to link the static files later in my project.
In the CMake documetation I saw you could use ExternalProject_Add() for the purpose of configuring and building an external project.
Just to test this I created a simple Project using CMake with ExternalProject_Add().
My test project has the following layout:
test
├── cmake-build-debug
│   ├── CMakeFiles
│   │   ├── 3.23.2
│   │   │   └── CompilerIdC
│   │   │   └── tmp
│   │   ├── CMakeTmp
│   │   ├── readline.dir
│   │   └── test.dir
│   ├── readline-prefix
│   │   ├── src
│   │   │   ├── readline-build
│   │   │   └── readline-stamp
│   │   └── tmp
│   └── Testing
│   └── Temporary
├── install_dir
└── libraries
└── readline-8.2
├── doc
├── examples
│   ├── autoconf
│   └── rlfe
├── m4
├── shlib
└── support
The library source files are located in test/libraries/readline-8.2.
I want the build files to get installed in the test/install_dir.
With the following Code I manage to build the the files, but I get errors when building it.
This is the content of my CMakeLists.txt
cmake_minimum_required(VERSION 3.23)
project(test C)
set(CMAKE_C_STANDARD 11)
set(LIB_PATH ${CMAKE_SOURCE_DIR}/libraries)
set(READLINE_LIB_SRC ${LIB_PATH}/readline-8.2)
set(CONFIGURE_ARGS --enable-shared=no --enable-install-examples=no --prefix=${CMAKE_SOURCE_DIR}/install_dir)
add_executable(test main.c)
include(ExternalProject)
ExternalProject_Add(
readline
BUILD_IN_SOURCE TRUE
BUILD_ALWAYS TRUE
SOURCE_DIR ${READLINE_LIB_SRC}
CONFIGURE_COMMAND ${READLINE_LIB_SRC}/configure ${CONFIGURE_ARGS}
BUILD_COMMAND make
INSTALL_COMMAND make install
)
I also tried adding the BUILD_BYPRODUCTS option of the ExternalProject_Add, as I read that ninja could face problems detecting targets otherwise, but without success.
The errors I get in the install step are:
mv: cannot stat '/home/devel/CLionProjects/test/install_dir/lib/libreadline.a': No such file or directory
mv: cannot stat '/home/devel/CLionProjects/test/install_dir/lib/libhistory.a': No such file or directory
But when I look at these directories the files exist.
The output of the build and the error message can be found here: https://gist.github.com/symb10sis/041f879ef5fbd02bcfd9200746101890
I couldn't include it in this question directly as it would get flagged as spam otherwise.
I spent hours reading the documetation, guides and questions here on stackoverflow but with my current knowledge of CMake I am stuck.
What causes the build to think these files don't exist? Is my use of ExternalProject_Add() appropriate for this or am I missing something?
I would really appreciate any help :)

How to read a raw signal from a microphone?

Linux Ubuntu 18.04
I'm trying to get the raw signal from a microphone as byte sequence and do some signal processing by myself. My first thought was actually just to find a device in the /dev directory, open file descriptor and start reading from it. But I discovered that there is audio library ALSA which I think does not really provide the ability to read a raw signal from the microphone.
So I found the /dev/snd directory and got confused. Here is how it looks like on my machine:
/dev/snd$ tree
.
├── by-path
│   ├── pci-0000:00:03.0 -> ../controlC0
│   └── pci-0000:00:1b.0 -> ../controlC1
├── controlC0
├── controlC1
├── hwC0D0
├── hwC1D0
├── pcmC0D3p
├── pcmC0D7p
├── pcmC0D8p
├── pcmC1D0c
├── pcmC1D0p
├── pcmC1D1p
├── pcmC1D2c
├── seq
└── timer
How to determine what is the microphone here?

Getting: conflicts with same file imported from GOPATH with google app engine, but not purely locally

I am still trying to deploy to google app engine both the local test environment and on app engine itself, and I keep getting this error:
2017/10/03 17:50:40 go-app-builder: Failed parsing input: app file DbConfig.go conflicts with same file imported from GOPATH
I have used grep to see if I explicitly import DbConfig but I, this is the result of grep:
$ grep -r 'DbConfig' .
./config.go:func GetDatabaseConfiguration() model.DbConfig {
./config.go: return model.DbConfig{connectionName, user, password, dbType, dbName}
Binary file ./share-services matches
./queries/db-config.go: dbConfig model.DbConfig
./queries/db-config.go:func Config(newDbConfig model.DbConfig) {
./queries/db-config.go: dbConfig = newDbConfig
./model/DbConfig.go:type DbConfig struct {
./model/DbConfig.go:func (d DbConfig) ConnectionString() string {
So from what I can tell I should not get an error and since other structs from my model package works just fine this is very strange to me. Firstly because Google App Engine accepts the deploy and first fails later, secondly because the dev server fails immediately, and lastly because if I run outside the app engine, I don't get this error.
Can someone give me a pointer to how to fix this please. I have search for a fix, but most state to use absolute include paths for packages which I already am.
EDIT:
I forgot to include the build command used, maybe that will help
go-app-builder -app_base /home/tools/go/src/bitbucket.com/chocolate-cloud-dev/share-services -arch 6 -dynamic -goroot /home/tools/sdks/google-cloud-sdk/platform/google_appengine/goroot-1.6 -gopath /home/tools/go -nobuild_files ^^$ -incremental_rebuild -unsafe -print_extras_hash links/generator.go model/user.go model/service.go model/SharedWith.go handlers.go config.go model/connection.go jsonHelpers/decoders.go queries/share-with-queries.go authentication/autenticate.go queries/share-queries.go model/share.go download/download.go model/Download.go model/File.go queries/file-queries.go jsonHelpers/encoders.go queries/db-config.go model/DbConfig.go model/NewestFile.go share-services.go jsonHelpers/CreateShare.go queries/download-queries.go
And this is my project structure:
.
├── app.yaml
├── authentication
│   └── autenticate.go
├── build-and-run
├── config.go
├── deply_v1.sh
├── download
│   └── download.go
├── go-notes.md
├── handlers.go
├── jsonHelpers
│   ├── CreateShare.go
│   ├── decoders.go
│   └── encoders.go
├── LICENSE
├── links
│   └── generator.go
├── model
│   ├── connection.go
│   ├── DbConfig.go
│   ├── Download.go
│   ├── File.go
│   ├── NewestFile.go
│   ├── service.go
│   ├── SharedWith.go
│   ├── share.go
│   └── user.go
├── models
├── queries
│   ├── db-config.go
│   ├── download-queries.go
│   ├── file-queries.go
│   ├── queries.sql
│   ├── share-queries.go
│   └── share-with-queries.go
├── README.md
├── run-locally-mac.sh
├── run-localy.sh
├── scripts
│   └── Script.sql
└── share-services.go
the answer is to either use relative paths instead of absolute, which still does not totaly work Or make a 100% flat folder structure.

Makefile header includes

I am developing a Makefile for an old project. The new Makefile is based on an older one I used successfully with other projects, but now I face an odd problem.
My project has the following form:
├── Makefile
└── src
├── folder1
│   ├── somefiles.c
│   └── somefiles.h
├── folder2
│   ├── problematic.h
│   ├── somefiles.c
│   ├── somefiles.h
│   └── subdir
│   └── included_file.h
└── folder3
├── somefiles.c
└── somefiles.h
My Makefile has the following line:
LIBRARY_INC_PATHS = $(addprefix -I, $(shell find . -name '*.h' -printf '"%h"\n' | sort -u))
In file problematic.h there is the line:
#include "subdir/included_file.h"
Which causes trouble. GCC says that there is no such file or directory.
If I change it to:
#include "included_file.h"
the problem vanishes, but this is not an acceptable solution in my case.
What modifications should I make to my makefile so I can include files from my C code using relative paths, instead of their name only?

Resources