blogc: common: moved "file" to common

4 files changed, 90 insertions, 4 deletions

diff --git a/src/common/error.c b/src/common/error.c
index 1ea6e0d..a6a5f24 100644
--- a/src/common/error.c
+++ b/src/common/error.c
@@ -112,6 +112,9 @@ bc_error_print(bc_error_t *err)
         case BC_ERROR_CONFIG_PARSER:
             fprintf(stderr, "error: config-parser: %s\n", err->msg);
             break;
+        case BC_ERROR_FILE:
+            fprintf(stderr, "error: file: %s\n", err->msg);
+            break;
         case BLOGC_ERROR_SOURCE_PARSER:
             fprintf(stderr, "error: source: %s\n", err->msg);
             break;
@@ -121,9 +124,6 @@ bc_error_print(bc_error_t *err)
         case BLOGC_ERROR_LOADER:
             fprintf(stderr, "error: loader: %s\n", err->msg);
             break;
-        case BLOGC_ERROR_FILE:
-            fprintf(stderr, "error: file: %s\n", err->msg);
-            break;
         case BLOGC_WARNING_DATETIME_PARSER:
             fprintf(stderr, "warning: datetime: %s\n", err->msg);
             break;
diff --git a/src/common/error.h b/src/common/error.h
index 2569538..d65a3e7 100644
--- a/src/common/error.h
+++ b/src/common/error.h
@@ -16,12 +16,12 @@ typedef enum {
 
     // errors for src/common
     BC_ERROR_CONFIG_PARSER = 1,
+    BC_ERROR_FILE,
 
     // errors for src/blogc
     BLOGC_ERROR_SOURCE_PARSER = 100,
     BLOGC_ERROR_TEMPLATE_PARSER,
     BLOGC_ERROR_LOADER,
-    BLOGC_ERROR_FILE,
     BLOGC_WARNING_DATETIME_PARSER,
 } bc_error_type_t;
 
diff --git a/src/common/file.c b/src/common/file.c
new file mode 100644
index 0000000..70a5631
--- /dev/null
+++ b/src/common/file.c
@@ -0,0 +1,67 @@
+/*
+ * blogc: A blog compiler.
+ * Copyright (C) 2015-2016 Rafael G. Martins <rafael@rafaelmartins.eng.br>
+ *
+ * This program can be distributed under the terms of the BSD License.
+ * See the file LICENSE.
+ */
+
+#include <errno.h>
+#include <stdarg.h>
+#include <stdbool.h>
+#include <stdint.h>
+#include <stdio.h>
+#include <string.h>
+#include "file.h"
+#include "error.h"
+#include "utf8.h"
+#include "utils.h"
+
+
+char*
+bc_file_get_contents(const char *path, size_t *len, bc_error_t **err)
+{
+    if (path == NULL || err == NULL || *err != NULL)
+        return NULL;
+
+    *len = 0;
+    FILE *fp = fopen(path, "r");
+
+    if (fp == NULL) {
+        int tmp_errno = errno;
+        *err = bc_error_new_printf(BC_ERROR_FILE,
+            "Failed to open file (%s): %s", path, strerror(tmp_errno));
+        return NULL;
+    }
+
+    bc_string_t *str = bc_string_new();
+    char buffer[BC_FILE_CHUNK_SIZE];
+    char *tmp;
+
+    while (!feof(fp)) {
+        size_t read_len = fread(buffer, sizeof(char), BC_FILE_CHUNK_SIZE, fp);
+
+        tmp = buffer;
+
+        if (str->len == 0 && read_len > 0) {
+            // skipping BOM before validation, for performance. should be safe
+            // enough
+            size_t skip = bc_utf8_skip_bom((uint8_t*) buffer, read_len);
+            read_len -= skip;
+            tmp += skip;
+        }
+
+        *len += read_len;
+        bc_string_append_len(str, tmp, read_len);
+    }
+    fclose(fp);
+
+    if (!bc_utf8_validate_str(str)) {
+        *err = bc_error_new_printf(BC_ERROR_FILE,
+            "File content is not valid UTF-8: %s", path);
+        bc_string_free(str, true);
+        return NULL;
+    }
+
+    return bc_string_free(str, false);
+}
diff --git a/src/common/file.h b/src/common/file.h
new file mode 100644
index 0000000..e095de7
--- /dev/null
+++ b/src/common/file.h
@@ -0,0 +1,19 @@
+/*
+ * blogc: A blog compiler.
+ * Copyright (C) 2015-2016 Rafael G. Martins <rafael@rafaelmartins.eng.br>
+ *
+ * This program can be distributed under the terms of the BSD License.
+ * See the file LICENSE.
+ */
+
+#ifndef _FILE_H
+#define _FILE_H
+
+#include <stddef.h>
+#include "error.h"
+
+#define BC_FILE_CHUNK_SIZE 1024
+
+char* bc_file_get_contents(const char *path, size_t *len, bc_error_t **err);
+
+#endif /* _FILE_H */