# Homeless Contents¶

This is a catch-all document for content that should be incorporated somewhere but for one reason or another has not yet found the right place.

## Mathematica behaviors / quirks¶

### Different behavior between the kernel and the notebook¶

The kernel and frontend have different parsers. The differences between them are usually semantically equivalent, but that is not always the case. This section collections meaningful differences between the two. See also Operator Precedence, which lists differences in operator precedence specifically.

#### x_. vs x_..¶

Notebook: The expression x_.. is a syntax error, because x_. is first parsed as Optional [Pattern[x,Blank[]]], and then . is encountered as an unexpected character.

Kernel: The .. is recognized as Repeated, so x_.. is parsed as Repeated[Pattern[x,Blank []]]

### Different behavior between ToExpression vs command line¶

The command line and ToExpression nearly always give the same output. Here are two examples where ToExpression does not match the command line.

Command line:

In[1]:= 2\:22127

Out[1]= 2


ToExpression:

In[1]:= ToExpression["2\\:" <> "22127"]

Out[1]= $Failed  On the command line, the following either emits a UTF-8 character encoding error and crashes or else results in the following (notice the skipped In/Out number): In[1]:= 2\[Minus]7 Out[3]= 2  ToExpression: In[1]:= ToExpression["2\\[" <> "Minus]7"] Out[1]=$Failed


### Frontend¶

#### RawEscape is as a letter¶

\[RawEscape] is supposed to be a letter-like character. Note also that \[RawEscape] is listed as a letter in UnicodeCharacters.tr. However, the FE has a hard time with it.

MathLinkCallFrontEnd[ FrontEndUndocumentedTestFEParserPacket["a\[RawEscape]a", False]]


The FE allows \[RawEscape] to begin a symbol name but not to end a symbol name. The FE interprets as \[RawEscape] at the end as implicit multiplication.

The command line does not have this problem, i.e. it treats \[RawEscape] as a letter without issue.