Add tab_parser

1 files changed, 78 insertions, 0 deletions

diff --git a/butl/tab-parser b/butl/tab-parser
new file mode 100644
index 0000000..6aa0705
--- /dev/null
+++ b/butl/tab-parser
@@ -0,0 +1,78 @@
+// file      : butl/tab-parser -*- C++ -*-
+// copyright : Copyright (c) 2014-2017 Code Synthesis Ltd
+// license   : MIT; see accompanying LICENSE file
+
+#ifndef BUTL_TAB_PARSER
+#define BUTL_TAB_PARSER
+
+#include <string>
+#include <vector>
+#include <iosfwd>
+#include <cstdint>   // uint64_t
+#include <stdexcept> // runtime_error
+
+#include <butl/export>
+
+#include <butl/char-scanner>
+
+namespace butl
+{
+  class LIBBUTL_EXPORT tab_parsing: public std::runtime_error
+  {
+  public:
+    tab_parsing (const std::string& name,
+                 std::uint64_t line,
+                 std::uint64_t column,
+                 const std::string& description);
+
+    std::string name;
+    std::uint64_t line;
+    std::uint64_t column;
+    std::string description;
+  };
+
+  // Line and columns are useful for issuing diagnostics about invalid or
+  // missing fields.
+  //
+  struct tab_field
+  {
+    std::string value;    // Field string (quoting preserved).
+    std::uint64_t column; // Field start column number (one-based).
+  };
+
+  struct tab_fields: std::vector<tab_field>
+  {
+    std::uint64_t line;       // Line number (one-based).
+    std::uint64_t end_column; // End-of-line column (line length).
+  };
+
+  // Read and parse lines consisting of space-separated fields. Field can
+  // contain single or double quoted substrings (with spaces) which are
+  // interpreted but preserved. No escaping of the quote characters is
+  // supported. Blank lines and lines that start with # (collectively called
+  // empty lines) are ignored.
+  //
+  class LIBBUTL_EXPORT tab_parser: protected butl::char_scanner
+  {
+  public:
+    tab_parser (std::istream& is, const std::string& name)
+        : char_scanner (is), name_ (name) {}
+
+    // Return next line of fields. Skip empty lines. Empty result denotes the
+    // end of stream.
+    //
+    tab_fields
+    next ();
+
+  private:
+    // Skip spaces and return the first peeked non-space character.
+    //
+    xchar
+    skip_spaces ();
+
+  private:
+    const std::string name_;
+  };
+}
+
+#endif // BUTL_TAB_PARSER