Merge pull request #13 from plos-clan/dev-copi143

修改magic的实现
plos-clan · Nov 6, 2024 · b0e8c5e · b0e8c5e
2 parents 683603b + 7d7a85e
commit b0e8c5e
Show file tree

Hide file tree

Showing 7 changed files with 515 additions and 777 deletions.
diff --git a/include/libc-base/string/str.h b/include/libc-base/string/str.h
@@ -84,18 +84,18 @@ OAPI char *strerror_r(int e, char *buf, size_t n) __THROW; // GNU 版本
 
 //; 自定义函数
 
-OAPI u32    utf8to32c(cstr8 *sp);
-OAPI size_t utf8to32s(u32 *d, cstr8 s);
-OAPI u32   *utf8to32a(cstr8 s);
-OAPI u32    utf16to32c(cstr16 *sp);
-OAPI size_t utf16to32s(u32 *d, cstr16 s);
-OAPI u32   *utf16to32a(cstr16 s);
-OAPI size_t utf32to8c(u32 c, u8 *s);
-OAPI size_t utf32to8s(u8 *d, cstr32 s);
-OAPI u8    *utf32to8a(cstr32 s);
-OAPI size_t utf32to16c(u32 c, u16 *s);
-OAPI size_t utf32to16s(u16 *d, cstr32 s);
-OAPI u16   *utf32to16a(cstr32 s);
+OAPI u32     utf8to32c(cstr8 *sp);
+OAPI size_t  utf8to32s(u32 *d, cstr8 s);
+OAPI u32    *utf8to32a(cstr8 s);
+OAPI u32     utf16to32c(cstr16 *sp);
+OAPI size_t  utf16to32s(u32 *d, cstr16 s);
+OAPI u32    *utf16to32a(cstr16 s);
+OAPI ssize_t utf32to8c(u32 c, u8 *s);
+OAPI size_t  utf32to8s(u8 *d, cstr32 s);
+OAPI u8     *utf32to8a(cstr32 s);
+OAPI ssize_t utf32to16c(u32 c, u16 *s);
+OAPI size_t  utf32to16s(u16 *d, cstr32 s);
+OAPI u16    *utf32to16a(cstr32 s);
 
 #undef IAPI
 #undef OAPI

diff --git a/include/magic.h b/include/magic.h
@@ -1,4 +1,26 @@
 #pragma once
 #include <define.h>
 
+enum {
+  MIME_UNKNOWN,
+  MIME_TEXT,
+  MIME_IMAGE,
+  MIME_AUDIO,
+  MIME_VIDEO,
+  MIME_APPLICATION,
+  MIME_FONT,
+  MIME_MESSAGE,
+  MIME_MODEL,
+  MIME_MULTIPART,
+};
+
+typedef struct filetype {
+  const i32  id;   // ID
+  const i32  type; // 类型
+  const cstr mime; // MIME 类型
+  const cstr desc; // 描述
+} *filetype_t;
+
 dlimport cstr filetype(const void *data, size_t size);
+
+dlexport cstr filetype_by_ext(cstr ext);
diff --git a/src/libc-base/string/charset.c b/src/libc-base/string/charset.c
@@ -37,7 +37,7 @@ u32 utf8to32c(cstr8 *sp) {
 }
 
 size_t utf8to32s(u32 *d, cstr8 s) {
-  u32 *_d = d;
+  const u32 *_d = d;
   while (*s != '\0') {
     *d++ = utf8to32c(&s);
   }
@@ -73,7 +73,7 @@ u32 utf16to32c(cstr16 *sp) {
 }
 
 size_t utf16to32s(u32 *d, cstr16 s) {
-  u32 *_d = d;
+  const u32 *_d = d;
   while (*s != '\0') {
     *d++ = utf16to32c(&s);
   }
@@ -88,7 +88,7 @@ u32 *utf16to32a(cstr16 s) {
   return realloc(r, (n + 1) * 4);
 }
 
-size_t utf32to8c(u32 c, u8 *s) {
+ssize_t utf32to8c(u32 c, u8 *s) {
   if (c <= 0x7f) {
     s[0] = c;
     return 1;
@@ -113,9 +113,9 @@ size_t utf32to8c(u32 c, u8 *s) {
 }
 
 size_t utf32to8s(u8 *d, cstr32 s) {
-  u8 *_d = d;
+  const u8 *_d = d;
   while (*s != '\0') {
-    size_t n = utf32to8c(*s++, d);
+    ssize_t n = utf32to8c(*s++, d);
     if (n < 0) {
       *d++ = 0xef;
       *d++ = 0xbf;
@@ -135,7 +135,7 @@ u8 *utf32to8a(cstr32 s) {
   return realloc(r, n + 1);
 }
 
-size_t utf32to16c(u32 c, u16 *s) {
+ssize_t utf32to16c(u32 c, u16 *s) {
   if (c <= 0xffff) {
     s[0] = c;
     return 1;
@@ -150,9 +150,9 @@ size_t utf32to16c(u32 c, u16 *s) {
 }
 
 size_t utf32to16s(u16 *d, cstr32 s) {
-  u16 *_d = d;
+  const u16 *_d = d;
   while (*s != '\0') {
-    size_t n = utf32to16c(*s++, d);
+    ssize_t n = utf32to16c(*s++, d);
     if (n < 0) {
       *d++ = 0xfffd;
     } else {
@@ -169,3 +169,76 @@ u16 *utf32to16a(cstr32 s) {
   size_t n = utf32to16s(r, s);
   return realloc(r, (n + 1) * 2);
 }
+
+bool is_vaild_utf32(u32 c) {
+  return c <= 0x10ffff && (c < 0xd800 || c > 0xdfff);
+}
+
+bool is_vaild_utf32s(cstr32 s) {
+  while (*s != '\0') {
+    if (!is_vaild_utf32(*s++)) return false;
+  }
+  return true;
+}
+
+size_t count_invalid_utf32s(cstr32 s) {
+  size_t n = 0;
+  while (*s != '\0') {
+    if (!is_vaild_utf32(*s++)) n++;
+  }
+  return n;
+}
+
+size_t remove_invalid_utf32s(u32 *d, cstr32 s) {
+  const u32 *_d = d;
+  while (*s != '\0') {
+    u32 c = *s++;
+    if (is_vaild_utf32(c)) *d++ = c;
+  }
+  *d = '\0';
+  return d - _d;
+}
+
+size_t count_invalid_utf16s(cstr16 s) {
+  size_t n = 0;
+  while (*s != '\0') {
+    u32 c = utf16to32c(&s);
+    if (!is_vaild_utf32(c)) n++;
+  }
+  return n;
+}
+
+size_t remove_invalid_utf16s(u16 *d, cstr16 s) {
+  const u16 *_d = d;
+  while (*s != '\0') {
+    u32 c = utf16to32c(&s);
+    if (is_vaild_utf32(c)) {
+      ssize_t n = utf32to16c(c, d);
+      if (n > 0) d += n;
+    }
+  }
+  *d = '\0';
+  return d - _d;
+}
+
+size_t count_invalid_utf8s(cstr8 s) {
+  size_t n = 0;
+  while (*s != '\0') {
+    u32 c = utf8to32c(&s);
+    if (!is_vaild_utf32(c)) n++;
+  }
+  return n;
+}
+
+size_t remove_invalid_utf8s(u8 *d, cstr8 s) {
+  const u8 *_d = d;
+  while (*s != '\0') {
+    u32 c = utf8to32c(&s);
+    if (is_vaild_utf32(c)) {
+      ssize_t n = utf32to8c(c, d);
+      if (n > 0) d += n;
+    }
+  }
+  *d = '\0';
+  return d - _d;
+}