Do not mark L mode JPEG as 1 bit in PDF

[radarhere]

Resolves #6149

When saving a PDF, the bits setting is different for 1 and L mode images.

Pillow/src/PIL/PdfImagePlugin.py

Lines 122 to 135 in a6a843e

	bits = 8
	params = None
	decode = None
	if im.mode == "1":
	filter = "DCTDecode"
	colorspace = PdfParser.PdfName("DeviceGray")
	procset = "ImageB" # grayscale
	bits = 1
	elif im.mode == "L":
	filter = "DCTDecode"
	# params = f"<< /Predictor 15 /Columns {width-2} >>"
	colorspace = PdfParser.PdfName("DeviceGray")
	procset = "ImageB" # grayscale

Except that both images are encoded as JPEGs,

Pillow/src/PIL/PdfImagePlugin.py

Lines 165 to 166 in a6a843e

	elif filter == "DCTDecode":
	Image.SAVE["JPEG"](im, op, filename)

which treats both 1 and L as L.

Pillow/src/PIL/JpegImagePlugin.py

Lines 574 to 578 in a6a843e

	# stuff to save JPEG files
	RAWMODE = {
	"1": "L",
	"L": "L",

So instructing the PDF that BitsPerComponent is different for these two images seems wrong. As a result, 1-bit PDFs from the current version of the code are read successfully by Chrome, but not by Firefox.

This PR fixes that by setting bits to 8 for 1-bit images as well.