Convert remaining sources to UTF-8 and comments to English

[omonien]

Hi Markus,

thanks for catching the umlaut breakage on the DUnitX PR (#97) — that ANSI→UTF-8 pass quietly turned prüfen into replacement junk. Sorry about that; you were right to flag it.

This follow-up does the rest of the tree in one clean pass, independent of #97:

• Remaining CP1252 sources → UTF-8 with BOM (so the IDE does not mis-detect ANSI again)
• Leftover German comments / docs → English
• Broken umlauts in DECHashBase.pas fixed by translating those NormanNG notes to English (not by stuffing more replacement characters in)
• Quick scan for other natural languages (French etc.) — none found beyond German

No crypto logic, no test vectors, no API changes. Pure encoding + human-language comments.

Scan report

Converted CP1252 → UTF-8 with BOM

• Source/DECCRC.pas
• Source/DECCiphers.pas
• Unit Tests/Tests/TestDECCipherModes.pas
• Unit Tests/Tests/TestDECCRC.pas
• Unit Tests/Tests/TestDECCipher.pas
• Unit Tests/Tests/TestDECCipherModesGCM.pas
• Unit Tests/Tests/TestDECCipherModesCCM.pas

UTF-8 BOM added to every other .pas / .dpr / .dpk / .inc that was already UTF-8 but missing the BOM (Delphi IDE convention). Existing BOMs left alone. No binaries / .res touched.

Comments translated (German → English), including

• Source/DECHashBase.pas — NormanNG Increment8 / EAX SELF notes (was full of U+FFFD)
• Source/DECCRC.pas, Source/DECCiphers.pas
• CCM / GCM “Additional Authentication Data prüfen” → “Verify additional authentication data”
• Various TODOs and inline notes in the cipher / hash / format / CRC unit tests
• Minor Germanism polynome → polynomial in CRC docs/messages
• Accidental ís → is in DECHash.pas

Other languages: none found (no French/Spanish/etc. comments or docs). Intentional umlauts in format test data (Decode('ä'), UTF-16 test strings) kept as real UTF-8 characters.

Take it whenever it suits you — should be easy to skim as encoding + comments only.

Cheers,
Olaf