feat: parse the CPU and motherboard model strings

Stripping by rule rather than by a table of known parts: a leading vendor
word, trademark noise, a trailing core count or clock. The Intel fixture
exists to keep it a rule.

DMI placeholders read as no board at all. 'To Be Filled By O.E.M.' on the
dashboard is worse than a blank row.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

4 files changed, 118 insertions, 0 deletions

diff --git a/lib/data.lua b/lib/data.lua
index 13e5ebd..1afc9c9 100644
--- a/lib/data.lua
+++ b/lib/data.lua
@@ -297,4 +297,69 @@ function M.kv_parse(text)
   return out
 end
 
+-- The CPU's marketing name, stripped to what identifies it.
+--
+-- '/proc/cpuinfo' repeats the model name once per thread; the first is enough.
+-- The stripping is rules, not a table of known CPUs: a leading vendor word,
+-- the registered-trademark noise, and a trailing core count or clock, all of
+-- which are constant boilerplate that costs a third of the row on a card
+-- measured in pixels.
+--
+--   AMD Ryzen 7 9700X 8-Core Processor      -> Ryzen 7 9700X
+--   Intel(R) Core(TM) i7-8700K CPU @ 3.7GHz -> Core i7-8700K
+function M.cpu_model(cpuinfo)
+  if type(cpuinfo) ~= 'string' then return nil end
+  local s = cpuinfo:match('model name%s*:%s*([^\n]+)')
+  if not s then return nil end
+  s = s:gsub('%(R%)', ''):gsub('%(TM%)', ''):gsub('%(tm%)', '')
+  s = s:gsub('^%s*AMD%s+', ''):gsub('^%s*Intel%s+', '')
+  s = s:gsub('%s+%d+%-Core Processor.*$', '')
+  s = s:gsub('%s+CPU%s*@.*$', '')
+  s = s:gsub('%s+Processor%s*$', '')
+  s = s:gsub('%s+', ' '):gsub('^%s+', ''):gsub('%s+$', '')
+  if s == '' then return nil end
+  return s
+end
+
+-- The motherboard, from the two world-readable DMI files.
+--
+-- board_vendor and board_name are readable without privilege, unlike the
+-- serial fields in the same directory. The vendor's corporate suffix is
+-- dropped because every vendor has one and none of it identifies the board.
+--
+-- A board reporting the DMI placeholder is treated as no board at all:
+-- 'To Be Filled By O.E.M.' on the dashboard is worse than a blank row.
+local DMI_PLACEHOLDER = {
+  ['to be filled by o.e.m.'] = true,
+  ['system manufacturer'] = true,
+  ['default string'] = true,
+  ['unknown'] = true,
+  ['n/a'] = true,
+}
+
+local function dmi_clean(s)
+  if type(s) ~= 'string' then return nil end
+  s = s:gsub('%s+', ' '):gsub('^%s+', ''):gsub('%s+$', '')
+  if s == '' or DMI_PLACEHOLDER[s:lower()] then return nil end
+  return s
+end
+
+function M.board_name(vendor, name)
+  vendor = dmi_clean(vendor)
+  name = dmi_clean(name)
+  if vendor then
+    -- Corporate boilerplate, longest first so 'Co., Ltd.' does not leave 'Co.'
+    vendor = vendor:gsub('%s+Technology Co%.,? Ltd%.?$', '')
+    vendor = vendor:gsub('%s+COMPUTER INC%.?$', '')
+    vendor = vendor:gsub('%s+Corporation$', ''):gsub('%s+Corp%.?$', '')
+    vendor = vendor:gsub('%s+Inc%.?$', ''):gsub('%s+INC%.?$', '')
+    vendor = vendor:gsub('%s+Co%.,?%s*Ltd%.?$', ''):gsub('%s+CO%.,?%s*LTD%.?$', '')
+    vendor = vendor:gsub('%s+GmbH$', ''):gsub('%s+LLC$', '')
+    vendor = vendor:gsub('%s+$', '')
+    if vendor == '' then vendor = nil end
+  end
+  if vendor and name then return vendor .. ' ' .. name end
+  return name or vendor
+end
+
 return M
diff --git a/test/fixtures/proc_cpuinfo b/test/fixtures/proc_cpuinfo
new file mode 100644
index 0000000..e9cd523
--- /dev/null
+++ b/test/fixtures/proc_cpuinfo
@@ -0,0 +1,12 @@
+processor	: 0
+vendor_id	: AuthenticAMD
+cpu family	: 26
+model		: 68
+model name	: AMD Ryzen 7 9700X 8-Core Processor
+stepping	: 0
+cpu MHz		: 4491.436
+cache size	: 1024 KB
+
+processor	: 1
+vendor_id	: AuthenticAMD
+model name	: AMD Ryzen 7 9700X 8-Core Processor
diff --git a/test/fixtures/proc_cpuinfo_intel b/test/fixtures/proc_cpuinfo_intel
new file mode 100644
index 0000000..886b589
--- /dev/null
+++ b/test/fixtures/proc_cpuinfo_intel
@@ -0,0 +1,6 @@
+processor	: 0
+vendor_id	: GenuineIntel
+cpu family	: 6
+model		: 158
+model name	: Intel(R) Core(TM) i7-8700K CPU @ 3.70GHz
+stepping	: 10
diff --git a/test/test_data.lua b/test/test_data.lua
index fdf0e2f..3e741e6 100644
--- a/test/test_data.lua
+++ b/test/test_data.lua
@@ -239,4 +239,39 @@ assert(type(data.kv_parse('')) == 'table', 'empty input gives a table')
 assert(type(data.kv_parse(nil)) == 'table', 'nil gives a table')
 assert(data.kv_parse('keyonly\n').keyonly == nil, 'a line with no value is skipped')
 
+-- === Hardware identification ==============================================
+-- Both strings are constant for the machine's life, so the widget memoizes
+-- them. What is tested here is the stripping, which must be a rule rather
+-- than a special case for one host: the Intel fixture exists to prove it.
+assert(data.cpu_model(read('test/fixtures/proc_cpuinfo')) == 'Ryzen 7 9700X',
+  'amd cpu, got ' .. tostring(data.cpu_model(read('test/fixtures/proc_cpuinfo'))))
+assert(data.cpu_model(read('test/fixtures/proc_cpuinfo_intel')) == 'Core i7-8700K',
+  'intel cpu, got ' .. tostring(data.cpu_model(read('test/fixtures/proc_cpuinfo_intel'))))
+
+-- Unparseable input yields nil, so the card shows nothing rather than a
+-- half-stripped string.
+assert(data.cpu_model('') == nil, 'empty cpuinfo gives nil')
+assert(data.cpu_model(nil) == nil, 'nil cpuinfo gives nil')
+assert(data.cpu_model('processor\t: 0\n') == nil, 'cpuinfo with no model name gives nil')
+
+-- The board is two sysfs files joined, with the vendor's corporate suffix
+-- dropped: it is boilerplate on every board and costs a third of the row.
+assert(data.board_name('Gigabyte Technology Co., Ltd.\n', 'X870 EAGLE WIFI7\n')
+  == 'Gigabyte X870 EAGLE WIFI7',
+  'board, got ' .. tostring(data.board_name('Gigabyte Technology Co., Ltd.\n', 'X870 EAGLE WIFI7\n')))
+assert(data.board_name('ASUSTeK COMPUTER INC.\n', 'PRIME B650-PLUS\n')
+  == 'ASUSTeK PRIME B650-PLUS',
+  'asus board, got ' .. tostring(data.board_name('ASUSTeK COMPUTER INC.\n', 'PRIME B650-PLUS\n')))
+
+-- A machine that reports one and not the other shows what it has.
+assert(data.board_name(nil, 'X870 EAGLE WIFI7\n') == 'X870 EAGLE WIFI7',
+  'name alone, got ' .. tostring(data.board_name(nil, 'X870 EAGLE WIFI7\n')))
+assert(data.board_name('Gigabyte\n', nil) == 'Gigabyte', 'vendor alone')
+assert(data.board_name(nil, nil) == nil, 'neither gives nil')
+
+-- A virtual machine reports placeholder DMI strings. Showing 'To be filled by
+-- O.E.M.' as the motherboard is worse than showing nothing.
+assert(data.board_name('To Be Filled By O.E.M.\n', 'To Be Filled By O.E.M.\n') == nil,
+  'placeholder DMI gives nil')
+
 print('test_data: all assertions passed')