First commit

conf/conf.go

@@ -0,0 +1,37 @@
+package conf
+
+import (
+	"bufio"
+	"errors"
+	"os"
+	"strings"
+)
+
+/* Object to handle what is in a JSON config */
+type Config struct {
+	configData map[string][]string;
+}
+
+func NewConfig(filePath string) (newConfig *Config, err error) {
+	// Open file
+	file, err := os.Open(filePath)
+	if err != nil {return}
+	defer file.Close()
+	
+	// Create configuration
+	newConfigData := make(map[string][]string);
+	
+	// Read lines into newConfigData
+	scanner := bufio.NewScanner(file)
+	for scanner.Scan() {
+	line := scanner.Text()
+		parts := strings.SplitN(line, " ", 2);
+		if len(parts) != 2 {err = errors.New("Bad config file!"); return;}
+		newConfigData[parts[0]] = append(newConfigData[parts[0]], parts[1]);
+	}
+	
+	// Create Config object and copy a reference to newConfigData into it
+	return &Config{configData: newConfigData}, nil;
+}
+
+func (config Config) GetConfigData() map[string][]string {return config.configData;}

conf/readme.md

@@ -0,0 +1,22 @@
+# Conf - Signal-CLI HTTP
+
+This module handles reading and parsing the config file, and acting as a verifier for the `Authorization` header on the HTTP requests.
+
+The config file is made up of multiple lines. The first token in each line is the `Authorization` bearer token. This cannot have spaces but can be any string. Choose wisely. The remainder of the line contains a path that the `Authorization` header is checked against. It does not matter if you include a leading or trailing slash.
+   
+Here's a sample config:
+
+```
+WGV99fSwgKhdQSa89HQIGxas /+16028675309/room/roomID/*
+WGV99fSwgKhdQSa89HQIGxas /+16028675309/direct/username.69/send
+ZQR3T6lqsvnXcgcWhpPOWWdv +16028675309/direct/username.69/send/
+```
+
+The config file is a **whitelist** for each bearer token to access a specific endpoint (or set of endpoints). The endpoints for this program are granular enough to only allow one action for each endpoint, so this level of whitelisting should™ be okay.
+
+There is a regex-like behavior to these paths using the `*` and `?` characters. For the regex-like behavior to be triggered these characters must be by themselves per path segment (no other characters not separated by a `/` or a start or end of string).
+
+The `*` character matches to any number of path segments. The `?` character matches to only one segment. Here's some examples:
+
+* `HZJWwB0TAjz6pjAHosII5ofR /+16028675309/*` will allow the bearer token to access any endpoint with the phone number `+16028675309`
+* `HZJWwB0TAjz6pjAHosII5ofR /+16028675309/direct/?/send` will allow the bearer token to send a direct message to anyone on that phone number.