I have created a list of all the characters that I have found for the TI-84+CSE (They probably work for other calculators). I think that source coder should recognize the real forms of the accented chars (ÁÀÂÄ) instead of nonsensical substitutes.

Here is the SourceCoder versions of all the versions
1234567890ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz
.,!?|!|?:;|#$%&@_|~'`...^^'^^`^^:
+-*/squareroot~xmarkintegral=invertedequal!=<<=>>=|<()[]{}
[|a][|b][|c][|d]|E[|e][e][|F]|F[i]|L|N[n][p]phat[phat][r][s][t]|u|v|w[xhat][yhat][z]
ÁÀÂÄáàâäÉÈÊËéèêëÍÌÎÏíìîïÓÒÔÖóòôöÚÙÛÜúùûüÇçÑñsharps
small0small1small2small3small4small5small6small7small8small9small10smallT^^T^^o^^r
[<|][|>]uparrowdownarrowbolduparrowbolddownarrowplotsquareplotcrossplotdotplotpixel
alphabetagammaDeltadeltaepsilonthetalambdamugreek_pipirhoSigmasigmatauPhichiOmega

Here is what is outputted to the calculator:
1234567890ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz
.,!?¡¿:;|#$%&@_~'`…´`¨
+-*/√-×∫==≠<≤>≥∠()[]{}
abcdEeeFFiʟNnpp̂p̂rstuvwx̄ȳz
ÁÀÂÄáàâäÉÈÊËéèêëÍÌÎÏíìîïÓÒÔÖóòôöÚÙÛÜúùûüÇçÑñß
◄►↑↓↑↓▫+⋅plotπxel
₀₁₂₃₄₅₆₇₈₉₁₀TT°r
αβγΔδελμππρΣστφχΩ

If anyone finds any that are missing then let me know. Does anyone know why there are multiple ways to write a character (e, [e], [|e])? I also noticed that plotpixel is outputted as plotπxel in sourcecoder.

The 'nonsensical characters' are used because most people don't want to have to dig around in their character map to access them for SourceCoder and Tokens. And most of the characters were mapped in TokenIDE first, and that's how he has them set up in the xml.