From 8d4b8f77d49110f04108eec6cbace6a33db78324 Mon Sep 17 00:00:00 2001
From: "me@ow.nekobit.net" <me@ow.nekobit.net>
Date: Fri, 8 Apr 2022 22:43:48 +0000
Subject: [PATCH] Mime parsing

FossilOrigin-Name: e6a37f930fb748cdf6828053bc6df089e4ea426602863ae77aefb62c8ff773f2
---
 .gitignore |   3 +-
 README.md  |   2 +
 src/mime.c | 112 +++++++++++++++++++++++++++++++++++++++++++----------
 src/mime.h |  11 +++++-
 4 files changed, 105 insertions(+), 23 deletions(-)

diff --git a/.gitignore b/.gitignore
index 3983865..a125849 100644
--- a/.gitignore
+++ b/.gitignore
@@ -4,4 +4,5 @@ filec
 **/*.chtml
 mastodont-c
 config.h
-treebird
\ No newline at end of file
+treebird
+test/tests
\ No newline at end of file
diff --git a/README.md b/README.md
index c357228..50ba5b6 100644
--- a/README.md
+++ b/README.md
@@ -1,3 +1,5 @@
+# NOTE: This software is not finished. Any bugs are likely noted
+
 # Treebird
 
 ![Treebird logo](./meta/treebird.png)
diff --git a/src/mime.c b/src/mime.c
index 0d9a6b2..fbb2301 100644
--- a/src/mime.c
+++ b/src/mime.c
@@ -20,6 +20,7 @@
 #include <fcgi_stdio.h>
 #include <string.h>
 #include <stdlib.h>
+#include <stddef.h>
 #include "mime.h"
 
 char* get_mime_boundary(char* content_type_str, char** bound)
@@ -64,49 +65,118 @@ error:
     return NULL;
 }
 
-char* strnws(char* str)
+static char* strnws(char* str)
 {
     for (; isblank(*str); ++str);
     return str;
 }
 
-#define STRSCMP(begin, str) strncmp((begin), (str), sizeof(str))
+// Function assumes character is at ';' char
+static char* read_key(char** r, char* key)
+{
+    char* val, *read_val;
+    size_t key_len = strlen(key);
+    *r = strnws(*r+1);
+    if (strncmp(*r, key, key_len) != 0)
+        return NULL;
+    *r = strnws(*r + key_len);
+    if (**r != '=')
+        return NULL;
+    *r = strnws(*r + 1);
+    if (**r == '\"') ++(*r);
+    val = *r;
+    // Chop off string
+    if ((read_val = strchr(*r, '\"')) ||
+        (read_val = strchr(*r, ';')) ||
+        (read_val = strchr(*r, '\r')))
+        *read_val = '\0';
+    // Jump to end
+    *r += strlen(val)+1;
+
+    return val;
+}
+
+#define STRSCMP(begin, str) strncmp((begin), (str), sizeof(str)-1)
+
 char* read_form_data(char* boundary, char* begin, struct http_form_info* info)
 {
-    char* r, *name;
+    ptrdiff_t data_size;
+    // Step 1: Parse name
+    // Parser variables
+    char* r, *read_val;
     size_t bound_len = strlen(boundary);
     if (strncmp(begin, "--", 2) != 0 ||
         strncmp(begin+2, boundary, bound_len) != 0)
         return NULL;
 
+    // Check if this is the end
+    if (strncmp(begin + 2 + bound_len, "--", 2) == 0)
+        return NULL;
+
     r = begin + 2 + bound_len + 2; /* Skip over LRSF */
 
     if (STRSCMP(r, "Content-Disposition") != 0)
         return NULL;
 
-    r = strnws(r + sizeof("Content-Disposition")+1);
-    if (*r != ':' ||
-        STRSCMP(r, "form-data") != 0) // Fucked up
+    r = strnws(r + sizeof("Content-Disposition")-1);
+    if (*r != ':') // Fucked up
         return NULL;
     
-    r = strnws(r + sizeof("form-data")+1);
+    r = strnws(r + 1);
+    if (STRSCMP(r, "form-data") != 0)
+        return NULL;
+    r = strnws(r + sizeof("form-data")-1);
     if (*r != ';')
         return NULL;
-    r = strnws(r+1);
-    if (STRSCMP(r, "name") != 0)
-        return NULL;
-    r = strnws(r + sizeof("name")+1);
-    if (*r != '=')
-        return NULL;
-    r = strnws(r + 1);
-    if (*r == '\"')
+    
+    info->name = read_key(&r, "name");
+    // Step 2: Parse filename (if there)
+    if (*r == ';')
+    {
+        info->filename = read_key(&r, "filename");
+    }
+    else
+        info->filename = NULL;
+
+    if (*r == '\r')
         r += 2;
-    if (*r == '\"') ++r;
-    name = r;
-    if ((r = strchr(r, '\"')))
-        *r = '\0';
 
-    printf("nameparsed: %s\r\n", name);
+    // Step 3: Parse Content-type (if there)
+    if (STRSCMP(r, "Content-Type") == 0)
+    {
+        r = strnws(r + sizeof("Content-type")-1);
+        if (*r != ':')
+            return NULL;
 
-    return name;
+        r = strnws(r + 1);
+        info->content_type = r;
+        if ((r = strchr(r, '\r')))
+            *r = '\0';
+        else
+            return NULL;
+        r += 4;
+    }
+    else if (*r == '\r') {
+        info->content_type = NULL;
+        r += 2;
+    }
+
+    // Last step: Find data
+    info->value = r;
+
+    // Look for end
+    if ((r = strstr(r, boundary)) &&
+        r && strstr(r-4, "\r\n--"))
+    {
+        r[-4] = '\0';
+        data_size = (r-4) - info->value;
+        info->value_size = data_size;
+    }
+    else
+        return NULL;
+
+    // Go back for next read
+    r -= 2;
+
+    return r;
 }
diff --git a/src/mime.h b/src/mime.h
index 631d79c..164f5bc 100644
--- a/src/mime.h
+++ b/src/mime.h
@@ -19,7 +19,16 @@
 #ifndef MIME_H
 #define MIME_H
 
+struct http_form_info
+{
+    char* name;
+    char* filename;
+    char* content_type;
+    char* value;
+    size_t value_size;
+};
+
 char* get_mime_boundary(char* content_type, char** res);
-char* read_mime_data(char* boundary, char* begin, struct http_form_info* info);
+char* read_form_data(char* boundary, char* begin, struct http_form_info* info);
 
 #endif /* MIME_H */