5 Lexical conventions [lex]

5.2 Phases of translation [lex.phases]

The precedence among the syntax rules of translation is specified by the following phases.7
An implementation shall support input files that are a sequence of UTF-8 code units (UTF-8 files).
It may also support an implementation-defined set of other kinds of input files, and, if so, the kind of an input file is determined in an implementation-defined manner that includes a means of designating input files as UTF-8 files, independent of their content.
[Note 1: 
In other words, recognizing the U+feff byte order mark is not sufficient.
— end note]
If an input file is determined to be a UTF-8 file, then it shall be a well-formed UTF-8 code unit sequence and it is decoded to produce a sequence of Unicode scalar values.
A sequence of translation character set elements is then formed by mapping each Unicode scalar value to the corresponding translation character set element.
In the resulting sequence, each pair of characters in the input sequence consisting of U+000d carriage return followed by U+000a line feed, as well as each U+000d carriage return not immediately followed by a U+000a line feed, is replaced by a single new-line character.
For any other kind of input file supported by the implementation, characters are mapped, in an implementation-defined manner, to a sequence of translation character set elements ([lex.charset]), representing end-of-line indicators as new-line characters.
If the first translation character is U+feff byte order mark, it is deleted.
Each sequence of a backslash character (\) immediately followed by zero or more whitespace characters other than new-line followed by a new-line character is deleted, splicing physical source lines to form logical source lines.
Only the last backslash on any physical source line shall be eligible for being part of such a splice.
[Note 2: 
Line splicing can form a universal-character-name ([lex.charset]).
— end note]
A source file that is not empty and that (after splicing) does not end in a new-line character shall be processed as if an additional new-line character were appended to the file.
The source file is decomposed into preprocessing tokens ([lex.pptoken]) and sequences of whitespace characters (including comments).
A source file shall not end in a partial preprocessing token or in a partial comment.8
Each comment is replaced by one space character.
New-line characters are retained.
Whether each nonempty sequence of whitespace characters other than new-line is retained or replaced by one space character is unspecified.
As characters from the source file are consumed to form the next preprocessing token (i.e., not being consumed as part of a comment or other forms of whitespace), except when matching a c-char-sequence, s-char-sequence, r-char-sequence, h-char-sequence, or q-char-sequence, universal-character-names are recognized and replaced by the designated element of the translation character set.
The process of dividing a source file's characters into preprocessing tokens is context-dependent.
[Example 1: 
See the handling of < within a #include preprocessing directive ([cpp.include]).
— end example]
Preprocessing directives are executed, macro invocations are expanded, and _Pragma unary operator expressions are executed.
A #include preprocessing directive causes the named header or source file to be processed from phase 1 through phase 4, recursively.
All preprocessing directives are then deleted.
For a sequence of two or more adjacent string-literal tokens, a common encoding-prefix is determined as specified in [lex.string].
Each such string-literal token is then considered to have that common encoding-prefix.
Adjacent string-literal tokens are concatenated ([lex.string]).
Whitespace characters separating tokens are no longer significant.
Each preprocessing token is converted into a token ([lex.token]).
The resulting tokens constitute a translation unit and are syntactically and semantically analyzed and translated.
[Note 3: 
The process of analyzing and translating the tokens can occasionally result in one token being replaced by a sequence of other tokens ([temp.names]).
— end note]
It is implementation-defined whether the sources for module units and header units on which the current translation unit has an interface dependency ([module.unit], [module.import]) are required to be available.
[Note 4: 
Source files, translation units and translated translation units need not necessarily be stored as files, nor need there be any one-to-one correspondence between these entities and any external representation.
The description is conceptual only, and does not specify any particular implementation.
— end note]
Translated translation units and instantiation units are combined as follows:
[Note 5: 
Some or all of these can be supplied from a library.
— end note]
Each translated translation unit is examined to produce a list of required instantiations.
[Note 6: 
This can include instantiations which have been explicitly requested ([temp.explicit]).
— end note]
The definitions of the required templates are located.
It is implementation-defined whether the source of the translation units containing these definitions is required to be available.
[Note 7: 
An implementation can choose to encode sufficient information into the translated translation unit so as to ensure the source is not required here.
— end note]
All the required instantiations are performed to produce instantiation units.
[Note 8: 
These are similar to translated translation units, but contain no references to uninstantiated templates and no template definitions.
— end note]
The program is ill-formed if any instantiation fails.
All external entity references are resolved.
Library components are linked to satisfy external references to entities not defined in the current translation.
All such translator output is collected into a program image which contains information needed for execution in its execution environment.
Implementations behave as if these separate phases occur, although in practice different phases can be folded together.
A partial preprocessing token would arise from a source file ending in the first portion of a multi-character token that requires a terminating sequence of characters, such as a header-name that is missing the closing " or >.
A partial comment would arise from a source file ending with an unclosed /* comment.